├── .gitignore ├── LICENSE ├── README.md ├── config ├── config.yaml ├── domainnet │ ├── C-S-1.json │ ├── C-S-3.json │ ├── P-C-1.json │ ├── P-C-3.json │ ├── P-R-1.json │ ├── P-R-3.json │ ├── R-C-1.json │ ├── R-C-3.json │ ├── R-P-1.json │ ├── R-P-3.json │ ├── R-S-1.json │ ├── R-S-3.json │ ├── S-P-1.json │ └── S-P-3.json ├── office │ ├── A-D-1.json │ ├── A-D-3.json │ ├── A-W-1.json │ ├── A-W-3.json │ ├── D-A-1.json │ ├── D-A-3.json │ ├── D-W-1.json │ ├── D-W-3.json │ ├── W-A-1.json │ ├── W-A-3.json │ ├── W-D-1.json │ └── W-D-3.json └── officehome │ ├── officehome_Ar-Cl-p03.json │ ├── officehome_Ar-Cl-p06.json │ ├── officehome_Ar-Pr-p03.json │ ├── officehome_Ar-Pr-p06.json │ ├── officehome_Ar-Rw-p03.json │ ├── officehome_Ar-Rw-p06.json │ ├── officehome_Cl-Ar-p03.json │ ├── officehome_Cl-Ar-p06.json │ ├── officehome_Cl-Pr-p03.json │ ├── officehome_Cl-Pr-p06.json │ ├── officehome_Cl-Rw-p03.json │ ├── officehome_Cl-Rw-p06.json │ ├── officehome_Pr-Ar-p03.json │ ├── officehome_Pr-Ar-p06.json │ ├── officehome_Pr-Cl-p03.json │ ├── officehome_Pr-Cl-p06.json │ ├── officehome_Pr-Rw-p03.json │ ├── officehome_Pr-Rw-p06.json │ ├── officehome_Rw-Ar-p03.json │ ├── officehome_Rw-Ar-p06.json │ ├── officehome_Rw-Cl-p03.json │ ├── officehome_Rw-Cl-p06.json │ ├── officehome_Rw-Pr-p03.json │ └── officehome_Rw-Pr-p06.json ├── data └── splits │ ├── domainnet │ ├── clipart.txt │ ├── clipart_labeled_1.txt │ ├── clipart_labeled_3.txt │ ├── clipart_unlabeled_1.txt │ ├── clipart_unlabeled_3.txt │ ├── painting.txt │ ├── painting_labeled_1.txt │ ├── painting_labeled_3.txt │ ├── painting_unlabeled_1.txt │ ├── painting_unlabeled_3.txt │ ├── real.txt │ ├── real_labeled_1.txt │ ├── real_labeled_3.txt │ ├── real_unlabeled_1.txt │ ├── real_unlabeled_3.txt │ ├── sketch.txt │ ├── sketch_labeled_1.txt │ ├── sketch_labeled_3.txt │ ├── sketch_unlabeled_1.txt │ └── sketch_unlabeled_3.txt │ ├── office │ ├── amazon.txt │ ├── amazon_labeled_1.txt │ ├── amazon_labeled_11.txt │ ├── amazon_labeled_3.txt │ ├── amazon_labeled_5.txt │ ├── amazon_labeled_7.txt │ ├── amazon_labeled_9.txt │ ├── amazon_unlabeled_1.txt │ ├── amazon_unlabeled_11.txt │ ├── amazon_unlabeled_3.txt │ ├── amazon_unlabeled_5.txt │ ├── amazon_unlabeled_7.txt │ ├── amazon_unlabeled_9.txt │ ├── dslr.txt │ ├── dslr_labeled_1.txt │ ├── dslr_labeled_11.txt │ ├── dslr_labeled_3.txt │ ├── dslr_labeled_5.txt │ ├── dslr_labeled_7.txt │ ├── dslr_labeled_9.txt │ ├── dslr_unlabeled_1.txt │ ├── dslr_unlabeled_11.txt │ ├── dslr_unlabeled_3.txt │ ├── dslr_unlabeled_5.txt │ ├── dslr_unlabeled_7.txt │ ├── dslr_unlabeled_9.txt │ ├── webcam.txt │ ├── webcam_labeled_1.txt │ ├── webcam_labeled_11.txt │ ├── webcam_labeled_3.txt │ ├── webcam_labeled_5.txt │ ├── webcam_labeled_7.txt │ ├── webcam_labeled_9.txt │ ├── webcam_unlabeled_1.txt │ ├── webcam_unlabeled_11.txt │ ├── webcam_unlabeled_3.txt │ ├── webcam_unlabeled_5.txt │ ├── webcam_unlabeled_7.txt │ └── webcam_unlabeled_9.txt │ ├── office_home │ ├── Art.txt │ ├── Art_labeled_p03.txt │ ├── Art_labeled_p06.txt │ ├── Art_unlabeled_p03.txt │ ├── Art_unlabeled_p06.txt │ ├── Clipart.txt │ ├── Clipart_labeled_p03.txt │ ├── Clipart_labeled_p06.txt │ ├── Clipart_unlabeled_p03.txt │ ├── Clipart_unlabeled_p06.txt │ ├── Product.txt │ ├── Product_labeled_p03.txt │ ├── Product_labeled_p06.txt │ ├── Product_unlabeled_p03.txt │ ├── Product_unlabeled_p06.txt │ ├── Real.txt │ ├── RealWorld.txt │ ├── RealWorld_labeled_p03.txt │ ├── RealWorld_labeled_p06.txt │ ├── RealWorld_unlabeled_p03.txt │ ├── RealWorld_unlabeled_p06.txt │ ├── Real_labeled_p03.txt │ ├── Real_labeled_p06.txt │ ├── Real_unlabeled_p03.txt │ └── Real_unlabeled_p06.txt │ └── visda17 │ ├── train.txt │ ├── train_labeled_p001.txt │ ├── train_labeled_p01.txt │ ├── train_unlabeled_p001.txt │ ├── train_unlabeled_p01.txt │ └── validation.txt ├── lib ├── iccv-poster.pdf ├── overview.png ├── result1.png └── result2.png ├── logs ├── domainnet │ ├── C-S-1.txt │ ├── C-S-3.txt │ ├── P-C-1.txt │ ├── P-C-3.txt │ ├── P-R-1.txt │ ├── P-R-3.txt │ ├── R-C-1.txt │ ├── R-C-3.txt │ ├── R-P-1.txt │ ├── R-P-3.txt │ ├── R-S-1.txt │ ├── R-S-3.txt │ ├── S-P-1.txt │ └── S-P-3.txt ├── office │ ├── A-D-1.txt │ ├── A-D-3.txt │ ├── A-W-1.txt │ ├── A-W-3.txt │ ├── D-A-1.txt │ ├── D-A-3.txt │ ├── D-W-1.txt │ ├── D-W-3.txt │ ├── W-A-1.txt │ ├── W-A-3.txt │ ├── W-D-1.txt │ └── W-D-3.txt └── officehome │ ├── ArCl-p03.txt │ ├── ArCl-p06.txt │ ├── ArPr-p03.txt │ ├── ArPr-p06.txt │ ├── ArRw-p03.txt │ ├── ArRw-p06.txt │ ├── ClAr-p03.txt │ ├── ClAr-p06.txt │ ├── ClPr-p03.txt │ ├── ClPr-p06.txt │ ├── ClRw-p03.txt │ ├── ClRw-p06.txt │ ├── PrAr-p03.txt │ ├── PrAr-p06.txt │ ├── PrCl-p03.txt │ ├── PrCl-p06.txt │ ├── PrRw-p03.txt │ ├── PrRw-p06.txt │ ├── RwAr-p03.txt │ ├── RwAr-p06.txt │ ├── RwCl-p03.txt │ ├── RwCl-p06.txt │ ├── RwPr-p03.txt │ └── RwPr-p06.txt ├── src ├── __init__.py ├── agents │ ├── BaseAgent.py │ ├── CVisDiT.py │ └── __init__.py ├── models │ ├── __init__.py │ ├── clustering.py │ ├── head.py │ ├── memorybank.py │ └── ssda.py ├── run.py └── utils │ ├── __init__.py │ ├── datautils.py │ ├── setup.py │ ├── torchutils.py │ └── utils.py └── train.sh /.gitignore: -------------------------------------------------------------------------------- 1 | # others 2 | exps/ 3 | 4 | # Created by https://www.toptal.com/developers/gitignore/api/python 5 | # Edit at https://www.toptal.com/developers/gitignore?templates=python 6 | 7 | ### Python ### 8 | # Byte-compiled / optimized / DLL files 9 | __pycache__/ 10 | *.py[cod] 11 | *$py.class 12 | 13 | # C extensions 14 | *.so 15 | 16 | # Distribution / packaging 17 | .Python 18 | build/ 19 | develop-eggs/ 20 | dist/ 21 | downloads/ 22 | eggs/ 23 | .eggs/ 24 | parts/ 25 | sdist/ 26 | var/ 27 | wheels/ 28 | pip-wheel-metadata/ 29 | share/python-wheels/ 30 | *.egg-info/ 31 | .installed.cfg 32 | *.egg 33 | MANIFEST 34 | 35 | # PyInstaller 36 | # Usually these files are written by a python script from a template 37 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 38 | *.manifest 39 | *.spec 40 | 41 | # Installer logs 42 | pip-log.txt 43 | pip-delete-this-directory.txt 44 | 45 | # Unit test / coverage reports 46 | htmlcov/ 47 | .tox/ 48 | .nox/ 49 | .coverage 50 | .coverage.* 51 | .cache 52 | nosetests.xml 53 | coverage.xml 54 | *.cover 55 | *.py,cover 56 | .hypothesis/ 57 | .pytest_cache/ 58 | pytestdebug.log 59 | 60 | # Translations 61 | *.mo 62 | *.pot 63 | 64 | # Django stuff: 65 | *.log 66 | local_settings.py 67 | db.sqlite3 68 | db.sqlite3-journal 69 | 70 | # Flask stuff: 71 | instance/ 72 | .webassets-cache 73 | 74 | # Scrapy stuff: 75 | .scrapy 76 | 77 | # Sphinx documentation 78 | docs/_build/ 79 | doc/_build/ 80 | 81 | # PyBuilder 82 | target/ 83 | 84 | # Jupyter Notebook 85 | .ipynb_checkpoints 86 | 87 | # IPython 88 | profile_default/ 89 | ipython_config.py 90 | 91 | # pyenv 92 | .python-version 93 | 94 | # pipenv 95 | # According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control. 96 | # However, in case of collaboration, if having platform-specific dependencies or dependencies 97 | # having no cross-platform support, pipenv may install dependencies that don't work, or not 98 | # install all needed dependencies. 99 | #Pipfile.lock 100 | 101 | # poetry 102 | #poetry.lock 103 | 104 | # PEP 582; used by e.g. github.com/David-OConnor/pyflow 105 | __pypackages__/ 106 | 107 | # Celery stuff 108 | celerybeat-schedule 109 | celerybeat.pid 110 | 111 | # SageMath parsed files 112 | *.sage.py 113 | 114 | # Environments 115 | # .env 116 | .env/ 117 | .venv/ 118 | env/ 119 | venv/ 120 | ENV/ 121 | env.bak/ 122 | venv.bak/ 123 | pythonenv* 124 | 125 | # Spyder project settings 126 | .spyderproject 127 | .spyproject 128 | 129 | # Rope project settings 130 | .ropeproject 131 | 132 | # mkdocs documentation 133 | /site 134 | 135 | # mypy 136 | .mypy_cache/ 137 | .dmypy.json 138 | dmypy.json 139 | 140 | # Pyre type checker 141 | .pyre/ 142 | 143 | # pytype static type analyzer 144 | .pytype/ 145 | 146 | # operating system-related files 147 | *.DS_Store #file properties cache/storage on macOS 148 | Thumbs.db #thumbnail cache on Windows 149 | 150 | # profiling data 151 | .prof 152 | 153 | # End of https://www.toptal.com/developers/gitignore/api/python 154 | 155 | # Created by https://www.toptal.com/developers/gitignore/api/vscode 156 | # Edit at https://www.toptal.com/developers/gitignore?templates=vscode 157 | 158 | ### vscode ### 159 | .vscode/* 160 | !.vscode/settings.json 161 | !.vscode/tasks.json 162 | !.vscode/launch.json 163 | !.vscode/extensions.json 164 | *.code-workspace 165 | 166 | # End of https://www.toptal.com/developers/gitignore/api/vscode 167 | 168 | # Created by https://www.toptal.com/developers/gitignore/api/macos 169 | # Edit at https://www.toptal.com/developers/gitignore?templates=macos 170 | 171 | ### macOS ### 172 | # General 173 | .DS_Store 174 | .AppleDouble 175 | .LSOverride 176 | 177 | # Icon must end with two \r 178 | Icon 179 | 180 | # Thumbnails 181 | ._* 182 | 183 | # Files that might appear in the root of a volume 184 | .DocumentRevisions-V100 185 | .fseventsd 186 | .Spotlight-V100 187 | .TemporaryItems 188 | .Trashes 189 | .VolumeIcon.icns 190 | .com.apple.timemachine.donotpresent 191 | 192 | # Directories potentially created on remote AFP share 193 | .AppleDB 194 | .AppleDesktop 195 | Network Trash Folder 196 | Temporary Items 197 | .apdisk 198 | 199 | # End of https://www.toptal.com/developers/gitignore/api/macos 200 | 201 | # Dataset 202 | data/visda17 203 | data/office 204 | data/officehome 205 | data/domainnet 206 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | MIT License 2 | 3 | Copyright (c) 2023 Bostoncake 4 | 5 | Permission is hereby granted, free of charge, to any person obtaining a copy 6 | of this software and associated documentation files (the "Software"), to deal 7 | in the Software without restriction, including without limitation the rights 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell 9 | copies of the Software, and to permit persons to whom the Software is 10 | furnished to do so, subject to the following conditions: 11 | 12 | The above copyright notice and this permission notice shall be included in all 13 | copies or substantial portions of the Software. 14 | 15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, 20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE 21 | SOFTWARE. 22 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Confidence-based Visual Dispersal for Few-shot Unsupervised Domain Adaptation (C-VisDiT) 2 | 3 | [![Paper](https://img.shields.io/badge/paper-arxiv.2309.15575-B31B1B.svg)](https://arxiv.org/pdf/2309.15575.pdf) 4 | [![Conference](https://img.shields.io/badge/iccv-2023-4b44ce.svg)](https://openaccess.thecvf.com/content/ICCV2023/html/Xiong_Confidence-based_Visual_Dispersal_for_Few-shot_Unsupervised_Domain_Adaptation_ICCV_2023_paper.html) 5 | 6 | 7 | Pytorch implementation of [C-VisDiT](https://openaccess.thecvf.com/content/ICCV2023/html/Xiong_Confidence-based_Visual_Dispersal_for_Few-shot_Unsupervised_Domain_Adaptation_ICCV_2023_paper.html) (**C**onfidence-based **Vis**ual **D**ispersal **T**ransfer) 8 | 9 | ## Overview 10 | 11 | ![C-VisDiT](./lib/overview.png) 12 | 13 | We present the Confidence-based Visual Dispersal Transfer learning method for Few-shot Unsupervised Domain Adaptation, aiming to comprehensively consider the importance of each sample during transfer based on its confidence. 14 | 15 | Compared with state-of-the-art methods, PCS improves the mean classification accuracy over different domain pairs on FUDA by **4.4%/1.7%** (1-shot/3-shots labeled source), **2.6%/2.8%** (3\%/6\% labeled source), **1.5%** (1\% labeled source), and **2.0%/2.5%** (1-shot/3-shots labeled source) on Office-31, Office-Home, VisDA-C, and DomainNet, respectively. 16 | 17 | ![result1](./lib/result1.png) 18 | 19 | ![result2](./lib/result2.png) 20 | 21 | ## Requirements 22 | 23 | ```bash 24 | conda create -n cvisdit python=3.7.11 25 | conda activate cvisdit 26 | 27 | conda install pytorch==1.8.0 torchvision==0.9.0 torchaudio==0.8.0 cudatoolkit=11.1 -c pytorch -c conda-forge 28 | pip install dotmap faiss-gpu==1.7.0 scikit-learn tensorboard tqdm 29 | ``` 30 | 31 | ## Training 32 | 33 | - Download the datasets from the Internet (Split files are provided in `data/splits`, we now support Office, Office-Home, VisDA-2017, and DomainNet) 34 | - Soft-link the datasets under the `data` folder (alias for corresponding datasets are `office`, `officehome`, `visda17`, and `domainnet`) 35 | - To train the model, please refer to `train.sh`. We provide training configurations and SOTA result training-logs in the `./config` and the `./logs` folders. Please note that model performance is not sensitive to hyper parameters added by C-VisDiT (`confidence_params`). These hyper parameters can change in a wide range without greatly affecting the final results. 36 | 37 | ## Citation 38 | 39 | ```bibtex 40 | @inproceedings{xiong2023confidence, 41 | title={Confidence-based Visual Dispersal for Few-shot Unsupervised Domain Adaptation}, 42 | author={Xiong, Yizhe and Chen, Hui and Lin, Zijia and Zhao, Sicheng and Ding, Guiguang}, 43 | booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision}, 44 | pages={11621--11631}, 45 | year={2023} 46 | } 47 | ``` 48 | 49 | ## Acknowlegdement 50 | 51 | This code is built on [[PCS](https://github.com/zhengzangw/PCS-FUDA)]. We thank the authors for sharing their code and some of the training configuration files. We reproduced some of the PCS results on our own. 52 | 53 | ## ToDo's 54 | 55 | - Upload training scripts for VisDA-C and BrAD-based DomainNet. 56 | -------------------------------------------------------------------------------- /config/config.yaml: -------------------------------------------------------------------------------- 1 | # This file illustrate keys' meaning in configs 2 | # This file is not a training config 3 | # training conig should be `xxx.json` 4 | # some default value are defined in `run.py:adjust_config` 5 | 6 | debug: false # Disable debug mode 7 | cuda: true # Use cuda 8 | gpu_device: null # use all available devices 9 | seed: 1337 # random seed 10 | exp_base: exps # directory that `experiments` folders will be created and experiment folder will be created under `{exp_base}/experiments/{exp_name}/{exp_id}` 11 | exp_name: office # experiment name 12 | exp_id: "1:D->A" # experiment id 13 | pretrained_exp_dir: null # folder where checkpoints can be loaded 14 | num_epochs: 500 # max number of epochs to run 15 | steps_epoch: 100 # number of iterations per epoch 16 | validate_freq: 1 # validation frequency 17 | copy_checkpoint_freq: 50 # frequency to copy the checkpoint 18 | data_params: 19 | name: office # name of dataset to use 20 | source: dslr # source domain 21 | target: amazon # target domain 22 | aug_src: aug_0 # augmentation for source 23 | aug_tgt: aug_0 # augmentation for target 24 | optim_params: 25 | learning_rate: 0.01 26 | conv_lr_ratio: 0.1 # ratio of learning for convolution layer 27 | patience: 4 # patience for early stop 28 | batch_size_lbd: 32 # batch size for labeled data 29 | batch_size: 64 30 | decay: true # use learning rate scheduler 31 | weight_decay: 5.e-4 32 | cls_update: true # update classifier's weight 33 | model_params: 34 | out_dim: 512 # feature dimension 35 | version: pretrain-resnet50 # network to use 36 | load_memory_bank: true 37 | # APCU hp 38 | load_weight: src-tgt 39 | load_weight_thres: 5 # threshold to load weight for one class 40 | load_weight_epoch: 5 # load after 5 epochs 41 | loss_params: 42 | temp: 0.1 # temparature for ssl 43 | thres_src: 0.99 # threshold for source psuedo (for APCU) 44 | thres_tgt: 0.99 45 | loss: [] # a list of loss 46 | # cls-so: supervised loss 47 | # proto-each + info: Loss InSelf 48 | # I2C-cross: Loss CrossSelf 49 | # semi-condentmax + semi-entmin: Loss MIM in source 50 | # tgt-condentmax + tgt-ent: Loss MIM in target 51 | weight: [] # weight for each loss 52 | clus: {} # clustering information 53 | -------------------------------------------------------------------------------- /config/domainnet/C-S-1.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "clipart->sketch:1", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "clipart", 8 | "target": "sketch", 9 | "fewshot": "1", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 5 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.5, 36 | 0.05, 37 | 0.5 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.001, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.001, 81 | "patience": 1 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.15, 86 | "lambda_mixed": 0.75, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.05, 89 | "target_inside_ratio": 0.03, 90 | "target_inside_confidence": 0.3 91 | }, 92 | "seed": 1337 93 | } -------------------------------------------------------------------------------- /config/domainnet/C-S-3.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "clipart->sketch:3", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "clipart", 8 | "target": "sketch", 9 | "fewshot": "3", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 5 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.5, 36 | 0.05, 37 | 0.5 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.01, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.0005, 81 | "patience": 4 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.3, 86 | "lambda_mixed": 1.0, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.05, 89 | "target_inside_ratio": 0.05, 90 | "target_inside_confidence": 0.35 91 | }, 92 | "seed": 1337 93 | } -------------------------------------------------------------------------------- /config/domainnet/P-C-1.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "painting->clipart:1", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "painting", 8 | "target": "clipart", 9 | "fewshot": "1", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 5 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.5, 36 | 0.05, 37 | 0.5 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.01, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.0001, 81 | "patience": 1 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.15, 86 | "lambda_mixed": 0.75, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.1, 89 | "target_inside_ratio": 0.03, 90 | "target_inside_confidence": 0.3 91 | }, 92 | "seed": 1337 93 | } -------------------------------------------------------------------------------- /config/domainnet/P-C-3.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "painting->clipart:3", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "painting", 8 | "target": "clipart", 9 | "fewshot": "3", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 5 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.5, 36 | 0.05, 37 | 0.5 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.01, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.0001, 81 | "patience": 4 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.15, 86 | "lambda_mixed": 1.25, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.15, 89 | "target_inside_ratio": 0.05, 90 | "target_inside_confidence": 0.5 91 | }, 92 | "seed": 1337 93 | } -------------------------------------------------------------------------------- /config/domainnet/P-R-1.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "painting->real:1", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "painting", 8 | "target": "real", 9 | "fewshot": "1", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 5 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.5, 36 | 0.05, 37 | 0.5 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.01, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.0005, 81 | "patience": 4 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.5, 86 | "lambda_mixed": 0.5, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.05, 89 | "target_inside_ratio": 0.15, 90 | "target_inside_confidence": 0.6 91 | }, 92 | "seed": 1337 93 | } -------------------------------------------------------------------------------- /config/domainnet/P-R-3.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "painting->real:3", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "painting", 8 | "target": "real", 9 | "fewshot": "3", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 5 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.5, 36 | 0.05, 37 | 0.5 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.01, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.0005, 81 | "patience": 4 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.5, 86 | "lambda_mixed": 0.7, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.1, 89 | "target_inside_ratio": 0.15, 90 | "target_inside_confidence": 0.6 91 | }, 92 | "seed": 1337 93 | } -------------------------------------------------------------------------------- /config/domainnet/R-C-1.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "real->clipart:1", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "real", 8 | "target": "clipart", 9 | "fewshot": "1", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 5 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.95, 36 | 0.05, 37 | 0.95 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.01, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.0005, 81 | "patience": 2 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.35, 86 | "lambda_mixed": 0.5, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.05, 89 | "target_inside_ratio": 0.05, 90 | "target_inside_confidence": 0.5 91 | }, 92 | "seed": 1337 93 | } -------------------------------------------------------------------------------- /config/domainnet/R-C-3.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "real->clipart:3", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "real", 8 | "target": "clipart", 9 | "fewshot": "3", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 5 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.5, 36 | 0.05, 37 | 0.5 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.01, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.0005, 81 | "patience": 4 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.35, 86 | "lambda_mixed": 1.0, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.1, 89 | "target_inside_ratio": 0.05, 90 | "target_inside_confidence": 0.5 91 | }, 92 | "seed": 1337 93 | } -------------------------------------------------------------------------------- /config/domainnet/R-P-1.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "real->painting:1", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "real", 8 | "target": "painting", 9 | "fewshot": "1", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 5 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.05, 36 | 0.05, 37 | 0.05 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.01, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.0005, 81 | "patience": 4 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.5, 86 | "lambda_mixed": 0.5, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.05, 89 | "target_inside_ratio": 0.15, 90 | "target_inside_confidence": 0.6 91 | }, 92 | "seed": 972056245 93 | } -------------------------------------------------------------------------------- /config/domainnet/R-P-3.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "real->painting:3", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "real", 8 | "target": "painting", 9 | "fewshot": "3", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 1 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.5, 36 | 0.05, 37 | 0.5 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.01, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.0005, 81 | "patience": 4 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.75, 86 | "lambda_mixed": 0.5, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.2, 89 | "target_inside_ratio": 0.25, 90 | "target_inside_confidence": 0.75 91 | }, 92 | "seed": 713214225 93 | } -------------------------------------------------------------------------------- /config/domainnet/R-S-1.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "real->sketch:1", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "real", 8 | "target": "sketch", 9 | "fewshot": "1", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 5 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.95, 36 | 0.05, 37 | 0.95 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.01, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.0005, 81 | "patience": 2 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.35, 86 | "lambda_mixed": 0.5, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.1, 89 | "target_inside_ratio": 0.05, 90 | "target_inside_confidence": 0.5 91 | }, 92 | "seed": 1583135841 93 | } -------------------------------------------------------------------------------- /config/domainnet/R-S-3.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "real->sketch:3", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "real", 8 | "target": "sketch", 9 | "fewshot": "3", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 2 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.05, 36 | 0.05, 37 | 0.05 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.01, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.0005, 81 | "patience": 4 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.35, 86 | "lambda_mixed": 1.25, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.2, 89 | "target_inside_ratio": 0.25, 90 | "target_inside_confidence": 0.75 91 | }, 92 | "seed": 1337 93 | } -------------------------------------------------------------------------------- /config/domainnet/S-P-1.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "sketch->painting:1", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "sketch", 8 | "target": "painting", 9 | "fewshot": "1", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 5 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.5, 36 | 0.05, 37 | 0.5 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.01, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.0005, 81 | "patience": 4 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.5, 86 | "lambda_mixed": 0.6, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.05, 89 | "target_inside_ratio": 0.15, 90 | "target_inside_confidence": 0.6 91 | }, 92 | "seed": 154484326 93 | } -------------------------------------------------------------------------------- /config/domainnet/S-P-3.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "domainnet", 4 | "exp_id": "sketch->painting:3", 5 | "data_params": { 6 | "name": "domainnet", 7 | "source": "sketch", 8 | "target": "painting", 9 | "fewshot": "3", 10 | "aug": "aug_0" 11 | }, 12 | "model_params": { 13 | "out_dim": 512, 14 | "version": "pretrain-resnet101", 15 | "load_memory_bank": true, 16 | "load_weight": "src-tgt", 17 | "load_weight_thres": 50, 18 | "load_weight_epoch": 5 19 | }, 20 | "loss_params": { 21 | "loss": [ 22 | "cls-so", 23 | "proto-each", 24 | "I2C-cross", 25 | "semi-entmin", 26 | "semi-condentmax", 27 | "tgt-entmin", 28 | "tgt-condentmax" 29 | ], 30 | "weight": [ 31 | 1, 32 | 1, 33 | 1, 34 | 0.05, 35 | 0.5, 36 | 0.05, 37 | 0.5 38 | ], 39 | "temp": 0.1, 40 | "clus": { 41 | "kmeans_freq": 4, 42 | "type": [ 43 | "each" 44 | ], 45 | "k": [ 46 | 126, 47 | 126, 48 | 126, 49 | 126, 50 | 126, 51 | 126, 52 | 126, 53 | 126, 54 | 126, 55 | 126, 56 | 252, 57 | 252, 58 | 252, 59 | 252, 60 | 252, 61 | 252, 62 | 252, 63 | 252, 64 | 252, 65 | 252 66 | ], 67 | "n_k": 1 68 | }, 69 | "thres_src": 0.99, 70 | "thres_tgt": 0.99 71 | }, 72 | "num_epochs": 500, 73 | "steps_epoch": null, 74 | "optim_params": { 75 | "batch_size_lbd": 32, 76 | "batch_size": 64, 77 | "learning_rate": 0.01, 78 | "conv_lr_ratio": 0.1, 79 | "decay": true, 80 | "weight_decay": 0.0005, 81 | "patience": 4 82 | }, 83 | "confidence_params": 84 | { 85 | "confidence_ratio": 0.5, 86 | "lambda_mixed": 0.7, 87 | "mixup_alpha": 0.75, 88 | "lambda_target_mixed": 0.1, 89 | "target_inside_ratio": 0.15, 90 | "target_inside_confidence": 0.6 91 | }, 92 | "seed": 1121162461 93 | } -------------------------------------------------------------------------------- /config/office/A-D-1.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "office", 4 | "exp_id": "amazon->dslr", 5 | "data_params": { 6 | "name": "office", 7 | "source": "amazon", 8 | "target": "dslr", 9 | "fewshot": "1", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 64, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 20, 29 | "load_weight_epoch": 2 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.98, 33 | "thres_tgt": 0.98, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [1, 1, 1, 0.5, 0.05, 0.5, 0.05], 45 | "clus": { 46 | "kmeans_freq": 1, 47 | "type": ["each"], 48 | "n_k": 15, 49 | "k": [31, 62] 50 | } 51 | }, 52 | "confidence_params": 53 | { 54 | "confidence_ratio": 0.95, 55 | "lambda_mixed": 1.0, 56 | "mixup_alpha": 0.75, 57 | "lambda_target_mixed": 0.25, 58 | "target_inside_ratio": 0.15, 59 | "target_inside_confidence": 0.75 60 | }, 61 | "seed": 1607667202 62 | } 63 | -------------------------------------------------------------------------------- /config/office/A-D-3.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "office", 4 | "exp_id": "amazon->dslr", 5 | "data_params": { 6 | "name": "office", 7 | "source": "amazon", 8 | "target": "dslr", 9 | "fewshot": "3", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 64, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 10, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.95, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [1, 1, 1, 0.1, 0.01, 0.1, 0.01], 45 | "clus": { 46 | "kmeans_freq": 1, 47 | "type": ["each"], 48 | "n_k": 10, 49 | "k": [31, 31, 62] 50 | } 51 | }, 52 | "confidence_params": 53 | { 54 | "confidence_ratio": 0.65, 55 | "lambda_mixed": 0.75, 56 | "mixup_alpha": 0.75, 57 | "lambda_target_mixed": 0.1, 58 | "target_inside_ratio": 0.2, 59 | "target_inside_confidence": 0.75 60 | }, 61 | "seed": 144705229 62 | } 63 | -------------------------------------------------------------------------------- /config/office/A-W-1.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "office", 4 | "exp_id": "amazon->webcam", 5 | "data_params": { 6 | "name": "office", 7 | "source": "amazon", 8 | "target": "webcam", 9 | "fewshot": "1", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 64, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 5, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.98, 33 | "thres_tgt": 0.98, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [1, 1, 1, 0.05, 0.05, 0.05, 0.05], 45 | "clus": { 46 | "kmeans_freq": 1, 47 | "type": ["each"], 48 | "n_k": 10, 49 | "k": [31, 31, 62] 50 | } 51 | }, 52 | "confidence_params": 53 | { 54 | "confidence_ratio": 0.5, 55 | "lambda_mixed": 0.5, 56 | "mixup_alpha": 0.75, 57 | "lambda_target_mixed": 0.1, 58 | "target_inside_ratio": 0.15, 59 | "target_inside_confidence": 0.75 60 | }, 61 | "seed": 527708456 62 | } 63 | -------------------------------------------------------------------------------- /config/office/A-W-3.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "office", 4 | "exp_id": "amazon->webcam", 5 | "data_params": { 6 | "name": "office", 7 | "source": "amazon", 8 | "target": "webcam", 9 | "fewshot": "3", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": 50, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 64, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 5, 29 | "load_weight_epoch": 1 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [1, 1, 1, 0.5, 0.05, 0.5, 0.05], 45 | "clus": { 46 | "kmeans_freq": 1, 47 | "type": ["each"], 48 | "n_k": 15, 49 | "k": [31, 62] 50 | } 51 | }, 52 | "confidence_params": 53 | { 54 | "confidence_ratio": 0.75, 55 | "lambda_mixed": 0.5, 56 | "mixup_alpha": 0.75, 57 | "lambda_target_mixed": 0.1, 58 | "target_inside_ratio": 0.05, 59 | "target_inside_confidence": 0.95 60 | }, 61 | "seed": 319277559 62 | } 63 | -------------------------------------------------------------------------------- /config/office/D-A-1.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "office", 4 | "exp_id": "dslr->amazon", 5 | "data_params": { 6 | "name": "office", 7 | "source": "dslr", 8 | "target": "amazon", 9 | "fewshot": "1", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 64, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 50, 29 | "load_weight_epoch": 3 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [1, 1, 1, 0.05, 0.05, 0.05, 0.05], 45 | "clus": { 46 | "kmeans_freq": 1, 47 | "type": ["each"], 48 | "n_k": 15, 49 | "k": [31, 62] 50 | } 51 | }, 52 | "confidence_params": 53 | { 54 | "confidence_ratio": 0.65, 55 | "lambda_mixed": 0.75, 56 | "mixup_alpha": 0.75, 57 | "lambda_target_mixed": 0.1, 58 | "target_inside_ratio": 0.2, 59 | "target_inside_confidence": 0.75 60 | }, 61 | "seed": 1605044524 62 | } 63 | -------------------------------------------------------------------------------- /config/office/D-A-3.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "office", 4 | "exp_id": "dslr->amazon", 5 | "data_params": { 6 | "name": "office", 7 | "source": "dslr", 8 | "target": "amazon", 9 | "fewshot": "3", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 64, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 50, 29 | "load_weight_epoch": 3 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [1, 1, 1, 0.05, 0.05, 0.05, 0.05], 45 | "clus": { 46 | "kmeans_freq": 1, 47 | "type": ["each"], 48 | "n_k": 15, 49 | "k": [31, 62] 50 | } 51 | }, 52 | "confidence_params": 53 | { 54 | "confidence_ratio": 0.65, 55 | "lambda_mixed": 0.5, 56 | "mixup_alpha": 0.75, 57 | "lambda_target_mixed": 0.25, 58 | "target_inside_ratio": 0.1, 59 | "target_inside_confidence": 0.8 60 | }, 61 | "seed": 1052987298 62 | } 63 | -------------------------------------------------------------------------------- /config/office/D-W-1.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "office", 4 | "exp_id": "dslr->webcam", 5 | "data_params": { 6 | "name": "office", 7 | "source": "dslr", 8 | "target": "webcam", 9 | "fewshot": "1", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 64, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 20, 29 | "load_weight_epoch": 0 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [1, 1, 1, 0.5, 0.05, 0.5, 0.05], 45 | "clus": { 46 | "kmeans_freq": 1, 47 | "type": ["each"], 48 | "n_k": 15, 49 | "k": [31, 62] 50 | } 51 | }, 52 | "confidence_params": 53 | { 54 | "confidence_ratio": 0.5, 55 | "lambda_mixed": 0.75, 56 | "mixup_alpha": 0.75, 57 | "lambda_target_mixed": 0.1, 58 | "target_inside_ratio": 0.15, 59 | "target_inside_confidence": 0.85 60 | }, 61 | "seed": 1063533579 62 | } 63 | -------------------------------------------------------------------------------- /config/office/D-W-3.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "office", 4 | "exp_id": "dslr->webcam", 5 | "data_params": { 6 | "name": "office", 7 | "source": "dslr", 8 | "target": "webcam", 9 | "fewshot": "3", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 64, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 20, 29 | "load_weight_epoch": 0 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [1, 1, 1, 0.05, 0.05, 0.05, 0.05], 45 | "clus": { 46 | "kmeans_freq": 1, 47 | "type": ["each"], 48 | "n_k": 15, 49 | "k": [31, 62] 50 | } 51 | }, 52 | "confidence_params": 53 | { 54 | "confidence_ratio": 0.65, 55 | "lambda_mixed": 1.5, 56 | "mixup_alpha": 0.75, 57 | "lambda_target_mixed": 0.1, 58 | "target_inside_ratio": 0.2, 59 | "target_inside_confidence": 0.85 60 | }, 61 | "seed": 111170657 62 | } 63 | -------------------------------------------------------------------------------- /config/office/W-A-1.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "office", 4 | "exp_id": "webcam->amazon", 5 | "data_params": { 6 | "name": "office", 7 | "source": "webcam", 8 | "target": "amazon", 9 | "fewshot": "1", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": 100, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 64, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 50, 29 | "load_weight_epoch": 2 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.98, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [1, 1, 1, 0.5, 0.05, 0.5, 0.05], 45 | "clus": { 46 | "kmeans_freq": 1, 47 | "type": ["each"], 48 | "n_k": 15, 49 | "k": [31, 62] 50 | } 51 | }, 52 | "confidence_params": 53 | { 54 | "confidence_ratio": 0.95, 55 | "lambda_mixed": 0.1, 56 | "mixup_alpha": 0.75, 57 | "lambda_target_mixed": 0.25, 58 | "target_inside_ratio": 0.1, 59 | "target_inside_confidence": 0.85 60 | }, 61 | "seed": 180768833 62 | } 63 | -------------------------------------------------------------------------------- /config/office/W-A-3.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "office", 4 | "exp_id": "webcam->amazon", 5 | "data_params": { 6 | "name": "office", 7 | "source": "webcam", 8 | "target": "amazon", 9 | "fewshot": "3", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 64, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 50, 29 | "load_weight_epoch": 3 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [1, 1, 1, 0.05, 0.05, 0.05, 0.05], 45 | "clus": { 46 | "kmeans_freq": 1, 47 | "type": ["each"], 48 | "n_k": 15, 49 | "k": [31, 62] 50 | } 51 | }, 52 | "confidence_params": 53 | { 54 | "confidence_ratio": 0.5, 55 | "lambda_mixed": 0.05, 56 | "mixup_alpha": 0.75, 57 | "lambda_target_mixed": 0.25, 58 | "target_inside_ratio": 0.1, 59 | "target_inside_confidence": 0.95 60 | }, 61 | "seed": 2025120063 62 | } 63 | -------------------------------------------------------------------------------- /config/office/W-D-1.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "office", 4 | "exp_id": "webcam->dslr", 5 | "data_params": { 6 | "name": "office", 7 | "source": "webcam", 8 | "target": "dslr", 9 | "fewshot": "1", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": 100, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 64, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 10, 29 | "load_weight_epoch": 0 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.95, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [1, 1, 1, 0.05, 0.05, 0.05, 0.05], 45 | "clus": { 46 | "kmeans_freq": 1, 47 | "type": ["each"], 48 | "n_k": 15, 49 | "k": [31, 62] 50 | } 51 | }, 52 | "confidence_params": 53 | { 54 | "confidence_ratio": 0.65, 55 | "lambda_mixed": 0.25, 56 | "mixup_alpha": 0.75, 57 | "lambda_target_mixed": 0.1, 58 | "target_inside_ratio": 0.05, 59 | "target_inside_confidence": 0.95 60 | }, 61 | "seed": 1723706915 62 | } 63 | -------------------------------------------------------------------------------- /config/office/W-D-3.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "office", 4 | "exp_id": "webcam->dslr", 5 | "data_params": { 6 | "name": "office", 7 | "source": "webcam", 8 | "target": "dslr", 9 | "fewshot": "3", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": 100, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 64, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 10, 29 | "load_weight_epoch": 0 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [1, 1, 1, 0.05, 0.05, 0.05, 0.05], 45 | "clus": { 46 | "kmeans_freq": 1, 47 | "type": ["each"], 48 | "n_k": 15, 49 | "k": [31, 62] 50 | } 51 | }, 52 | "confidence_params": 53 | { 54 | "confidence_ratio": 0.95, 55 | "lambda_mixed": 0.25, 56 | "mixup_alpha": 0.75, 57 | "lambda_target_mixed": 0.1, 58 | "target_inside_ratio": 0.2, 59 | "target_inside_confidence": 0.75 60 | }, 61 | "seed": 7646520 62 | } 63 | -------------------------------------------------------------------------------- /config/officehome/officehome_Ar-Cl-p03.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Art->Clipart:p03", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Art", 8 | "target": "Clipart", 9 | "fewshot": "p03", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 30, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.95, 33 | "thres_tgt": 0.98, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-src", 38 | "proto-tgt", 39 | "I2C-cross", 40 | "semi-condentmax", 41 | "semi-entmin", 42 | "tgt-condentmax", 43 | "tgt-entmin" 44 | ], 45 | "weight": [1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 46 | "clus": { 47 | "kmeans_freq": 1, 48 | "type": ["each"], 49 | "n_k": 15, 50 | "k": [65, 130] 51 | } 52 | }, 53 | "confidence_params": 54 | { 55 | "confidence_ratio": 0.75, 56 | "lambda_mixed": 1.0, 57 | "mixup_alpha": 0.75, 58 | "lambda_target_mixed": 0.2, 59 | "target_inside_ratio": 0.25, 60 | "target_inside_confidence": 0.75 61 | }, 62 | "seed": 1337 63 | } 64 | -------------------------------------------------------------------------------- /config/officehome/officehome_Ar-Cl-p06.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Art->Clipart:p06", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Art", 8 | "target": "Clipart", 9 | "fewshot": "p06", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 30, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.95, 33 | "thres_tgt": 0.98, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-src", 38 | "proto-tgt", 39 | "I2C-cross", 40 | "semi-condentmax", 41 | "semi-entmin", 42 | "tgt-condentmax", 43 | "tgt-entmin" 44 | ], 45 | "weight": [1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 46 | "clus": { 47 | "kmeans_freq": 1, 48 | "type": ["each"], 49 | "n_k": 15, 50 | "k": [65, 130] 51 | } 52 | }, 53 | "confidence_params": 54 | { 55 | "confidence_ratio": 0.25, 56 | "lambda_mixed": 1.0, 57 | "mixup_alpha": 0.75, 58 | "lambda_target_mixed": 0.1, 59 | "target_inside_ratio": 0.15, 60 | "target_inside_confidence": 0.75 61 | }, 62 | "seed": 1337 63 | } 64 | -------------------------------------------------------------------------------- /config/officehome/officehome_Ar-Pr-p03.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Art->Product:p03", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Art", 8 | "target": "Product", 9 | "fewshot": "p03", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 30 29 | }, 30 | "loss_params": { 31 | "thres_src": 0.99, 32 | "thres_tgt": 0.99, 33 | "temp": 0.1, 34 | "loss": [ 35 | "cls-so", 36 | "proto-each", 37 | "I2C-cross", 38 | "semi-condentmax", 39 | "semi-entmin", 40 | "tgt-condentmax", 41 | "tgt-entmin" 42 | ], 43 | "weight": [ 44 | 1, 45 | 1, 46 | 1, 47 | 0.05, 48 | 0.05, 49 | 0.05, 50 | 0.05 51 | ], 52 | "clus": { 53 | "kmeans_freq": 1, 54 | "type": ["each"], 55 | "n_k": 10, 56 | "k": [65, 130, 130] 57 | } 58 | }, 59 | "confidence_params": 60 | { 61 | "confidence_ratio": 0.75, 62 | "lambda_mixed": 1.0, 63 | "mixup_alpha": 0.75, 64 | "lambda_target_mixed": 0.1, 65 | "target_inside_ratio": 0.2, 66 | "target_inside_confidence": 0.75 67 | }, 68 | "seed": 1850702017 69 | } 70 | -------------------------------------------------------------------------------- /config/officehome/officehome_Ar-Pr-p06.json: -------------------------------------------------------------------------------- 1 | { 2 | "debug": false, 3 | "cuda": true, 4 | "gpu_device": null, 5 | "exp_base": "./exps", 6 | "exp_name": "officehome", 7 | "exp_id": "Art->Product:p06", 8 | "pretrained_exp_dir": null, 9 | "data_params": { 10 | "name": "office_home", 11 | "source": "Art", 12 | "aug_src": "aug_0", 13 | "target": "Product", 14 | "aug_tgt": "aug_0", 15 | "fewshot": "p06", 16 | "train_val_split": false, 17 | "image_size": 224 18 | }, 19 | "model_params": { 20 | "out_dim": 512, 21 | "version": "pretrain-resnet50", 22 | "load_memory_bank": true, 23 | "load_weight": "src-tgt", 24 | "load_weight_thres": 50, 25 | "cls_update": true 26 | }, 27 | "loss_params": { 28 | "loss": [ 29 | "cls-so", 30 | "proto-each", 31 | "I2C-cross", 32 | "semi-condentmax", 33 | "semi-entmin", 34 | "tgt-condentmax", 35 | "tgt-entmin" 36 | ], 37 | "weight": [ 38 | 1, 39 | 1, 40 | 1, 41 | 0.05, 42 | 0.05, 43 | 0.05, 44 | 0.05 45 | ], 46 | "start": null, 47 | "end": null, 48 | "temp": 0.1, 49 | "m": 0.5, 50 | "T": 0.05, 51 | "clus": { 52 | "kmeans_freq": 1, 53 | "type": [ 54 | "each" 55 | ], 56 | "k": [ 57 | 65, 58 | 65, 59 | 65, 60 | 65, 61 | 65, 62 | 65, 63 | 65, 64 | 65, 65 | 65, 66 | 65, 67 | 130, 68 | 130, 69 | 130, 70 | 130, 71 | 130, 72 | 130, 73 | 130, 74 | 130, 75 | 130, 76 | 130, 77 | 130, 78 | 130, 79 | 130, 80 | 130, 81 | 130, 82 | 130, 83 | 130, 84 | 130, 85 | 130, 86 | 130 87 | ], 88 | "n_k":1 89 | }, 90 | "pseudo": true, 91 | "thres_src": 0.99, 92 | "thres_tgt": 0.99 93 | }, 94 | "num_epochs": 500, 95 | "steps_epoch": null, 96 | "optim_params": { 97 | "batch_size_lbd": 32, 98 | "batch_size_src": 64, 99 | "batch_size_tgt": 64, 100 | "batch_size": 64, 101 | "combine": false, 102 | "learning_rate": 0.01, 103 | "conv_lr_ratio": 0.1, 104 | "momentum": 0.9, 105 | "nesterov": true, 106 | "decay": true, 107 | "weight_decay": 0.0005, 108 | "patience": 4 109 | }, 110 | "validate_freq": 1, 111 | "visualize": { 112 | "feature_analysis_freq": 1, 113 | "log": true, 114 | "stat": true, 115 | "svm_freq": null, 116 | "tsne_freq": null, 117 | "tsne_clus": false, 118 | "clus_freq": null, 119 | "confusion_freq": null, 120 | "error_freq": null 121 | }, 122 | "copy_checkpoint_freq": 20, 123 | "confidence_params": 124 | { 125 | "confidence_ratio": 0.75, 126 | "lambda_mixed": 1.0, 127 | "mixup_alpha": 0.75, 128 | "lambda_target_mixed": 0.1, 129 | "target_inside_ratio": 0.01, 130 | "target_inside_confidence": 0.95 131 | }, 132 | "seed": 811088573 133 | } -------------------------------------------------------------------------------- /config/officehome/officehome_Ar-Rw-p03.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Art->Real:p03", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Art", 8 | "target": "Real", 9 | "fewshot": "p03", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 30, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.95, 33 | "thres_tgt": 0.98, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-src", 38 | "proto-tgt", 39 | "I2C-cross", 40 | "semi-condentmax", 41 | "semi-entmin", 42 | "tgt-condentmax", 43 | "tgt-entmin" 44 | ], 45 | "weight": [1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 46 | "clus": { 47 | "kmeans_freq": 1, 48 | "type": ["each"], 49 | "n_k": 15, 50 | "k": [65, 130] 51 | } 52 | }, 53 | "confidence_params": 54 | { 55 | "confidence_ratio": 0.75, 56 | "lambda_mixed": 1.0, 57 | "mixup_alpha": 0.75, 58 | "lambda_target_mixed": 0.2, 59 | "target_inside_ratio": 0.15, 60 | "target_inside_confidence": 0.75 61 | }, 62 | "seed": 1337 63 | } 64 | -------------------------------------------------------------------------------- /config/officehome/officehome_Ar-Rw-p06.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Art->Real:p06", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Art", 8 | "target": "Real", 9 | "fewshot": "p06", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 30, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.95, 33 | "thres_tgt": 0.98, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-src", 38 | "proto-tgt", 39 | "I2C-cross", 40 | "semi-condentmax", 41 | "semi-entmin", 42 | "tgt-condentmax", 43 | "tgt-entmin" 44 | ], 45 | "weight": [1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 46 | "clus": { 47 | "kmeans_freq": 1, 48 | "type": ["each"], 49 | "n_k": 15, 50 | "k": [65, 130] 51 | } 52 | }, 53 | "confidence_params": 54 | { 55 | "confidence_ratio": 0.75, 56 | "lambda_mixed": 1.0, 57 | "mixup_alpha": 0.75, 58 | "lambda_target_mixed": 0.1, 59 | "target_inside_ratio": 0.15, 60 | "target_inside_confidence": 0.75 61 | }, 62 | "seed": 1919164751 63 | } 64 | -------------------------------------------------------------------------------- /config/officehome/officehome_Cl-Ar-p03.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Clipart->Art:p03", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Clipart", 8 | "target": "Art", 9 | "fewshot": "p03", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 10 29 | }, 30 | "loss_params": { 31 | "thres_src": 0.95, 32 | "thres_tgt": 0.99, 33 | "temp": 0.1, 34 | "loss": [ 35 | "cls-so", 36 | "proto-src", 37 | "proto-tgt", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [ 45 | 1, 46 | 1, 47 | 0.5, 48 | 1, 49 | 0.5, 50 | 0.05, 51 | 0.5, 52 | 0.05 53 | ], 54 | "clus": { 55 | "kmeans_freq": 1, 56 | "type": ["each"], 57 | "n_k": 10, 58 | "k": [65, 130, 130] 59 | } 60 | }, 61 | "confidence_params": 62 | { 63 | "confidence_ratio": 0.75, 64 | "lambda_mixed": 1.0, 65 | "mixup_alpha": 0.75, 66 | "lambda_target_mixed": 0.1, 67 | "target_inside_ratio": 0.15, 68 | "target_inside_confidence": 0.75 69 | }, 70 | "seed": 1730196238 71 | } 72 | -------------------------------------------------------------------------------- /config/officehome/officehome_Cl-Ar-p06.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Clipart->Art:p06", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Clipart", 8 | "target": "Art", 9 | "fewshot": "p06", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 50 29 | }, 30 | "loss_params": { 31 | "thres_src": 0.99, 32 | "thres_tgt": 0.99, 33 | "temp": 0.1, 34 | "loss": [ 35 | "cls-so", 36 | "proto-each", 37 | "I2C-cross", 38 | "semi-condentmax", 39 | "semi-entmin", 40 | "tgt-condentmax", 41 | "tgt-entmin" 42 | ], 43 | "weight": [ 44 | 1, 45 | 1, 46 | 1, 47 | 0.05, 48 | 0.05, 49 | 0.05, 50 | 0.05 51 | ], 52 | "clus": { 53 | "kmeans_freq": 1, 54 | "type": ["each"], 55 | "n_k": 10, 56 | "k": [65, 130, 130] 57 | } 58 | }, 59 | "confidence_params": 60 | { 61 | "confidence_ratio": 0.75, 62 | "lambda_mixed": 1.0, 63 | "mixup_alpha": 0.75, 64 | "lambda_target_mixed": 0.1, 65 | "target_inside_ratio": 0.1, 66 | "target_inside_confidence": 0.95 67 | }, 68 | "seed": 1498011406 69 | } 70 | -------------------------------------------------------------------------------- /config/officehome/officehome_Cl-Pr-p03.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Clipart->Product:p03", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Clipart", 8 | "target": "Product", 9 | "fewshot": "p03", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 50, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [ 45 | 1, 46 | 1, 47 | 1, 48 | 0.05, 49 | 0.05, 50 | 0.05, 51 | 0.05 52 | ], 53 | "clus": { 54 | "kmeans_freq": 1, 55 | "type": ["each"], 56 | "n_k": 10, 57 | "k": [65, 130, 130] 58 | } 59 | }, 60 | "confidence_params": 61 | { 62 | "confidence_ratio": 0.75, 63 | "lambda_mixed": 1.0, 64 | "mixup_alpha": 0.75, 65 | "lambda_target_mixed": 0.1, 66 | "target_inside_ratio": 0.25, 67 | "target_inside_confidence": 0.95 68 | }, 69 | "seed": 1462296150 70 | } 71 | -------------------------------------------------------------------------------- /config/officehome/officehome_Cl-Pr-p06.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Clipart->Product:p06", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Clipart", 8 | "target": "Product", 9 | "fewshot": "p06", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 50, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [ 45 | 1, 46 | 1, 47 | 1, 48 | 0.05, 49 | 0.05, 50 | 0.05, 51 | 0.05 52 | ], 53 | "clus": { 54 | "kmeans_freq": 1, 55 | "type": ["each"], 56 | "n_k": 10, 57 | "k": [65, 130, 130] 58 | } 59 | }, 60 | "confidence_params": 61 | { 62 | "confidence_ratio": 0.75, 63 | "lambda_mixed": 1.0, 64 | "mixup_alpha": 0.75, 65 | "lambda_target_mixed": 0.2, 66 | "target_inside_ratio": 0.1, 67 | "target_inside_confidence": 0.75 68 | }, 69 | "seed": 1871350851 70 | } 71 | -------------------------------------------------------------------------------- /config/officehome/officehome_Cl-Rw-p03.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Clipart->Real:p03", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Clipart", 8 | "target": "Real", 9 | "fewshot": "p03", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 50, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-each", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [ 45 | 1, 46 | 1, 47 | 1, 48 | 0.05, 49 | 0.05, 50 | 0.05, 51 | 0.05 52 | ], 53 | "clus": { 54 | "kmeans_freq": 1, 55 | "type": ["each"], 56 | "n_k": 10, 57 | "k": [65, 130, 130] 58 | } 59 | }, 60 | "confidence_params": 61 | { 62 | "confidence_ratio": 0.75, 63 | "lambda_mixed": 1.0, 64 | "mixup_alpha": 0.75, 65 | "lambda_target_mixed": 0.1, 66 | "target_inside_ratio": 0.02, 67 | "target_inside_confidence": 0.95 68 | }, 69 | "seed": 1700010695 70 | } 71 | -------------------------------------------------------------------------------- /config/officehome/officehome_Cl-Rw-p06.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Clipart->Real:p06", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Clipart", 8 | "target": "Real", 9 | "fewshot": "p06", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 10, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "cls-info", 38 | "proto-src", 39 | "proto-tgt", 40 | "I2C-cross", 41 | "semi-condentmax", 42 | "semi-entmin", 43 | "tgt-condentmax", 44 | "tgt-entmin" 45 | ], 46 | "weight": [1, 0.1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 47 | "clus": { 48 | "kmeans_freq": 1, 49 | "type": ["each"], 50 | "n_k": 15, 51 | "k": [65, 130] 52 | } 53 | }, 54 | "confidence_params": 55 | { 56 | "confidence_ratio": 0.75, 57 | "lambda_mixed": 1.0, 58 | "mixup_alpha": 0.75, 59 | "lambda_target_mixed": 0.1, 60 | "target_inside_ratio": 0.03, 61 | "target_inside_confidence": 0.95 62 | }, 63 | "seed": 951628884 64 | } 65 | -------------------------------------------------------------------------------- /config/officehome/officehome_Pr-Ar-p03.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Product->Art:p03", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Product", 8 | "target": "Art", 9 | "fewshot": "p03", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 10, 29 | "load_weight_epoch": 0 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.95, 33 | "thres_tgt": 0.98, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-src", 38 | "proto-tgt", 39 | "I2C-cross", 40 | "semi-condentmax", 41 | "semi-entmin", 42 | "tgt-condentmax", 43 | "tgt-entmin" 44 | ], 45 | "weight": [1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 46 | "clus": { 47 | "kmeans_freq": 1, 48 | "type": ["each"], 49 | "n_k": 15, 50 | "k": [65, 130] 51 | } 52 | }, 53 | "confidence_params": 54 | { 55 | "confidence_ratio": 0.75, 56 | "lambda_mixed": 1.0, 57 | "mixup_alpha": 0.75, 58 | "lambda_target_mixed": 0.1, 59 | "target_inside_ratio": 0.1, 60 | "target_inside_confidence": 0.95 61 | }, 62 | "seed": 1887460539 63 | } 64 | -------------------------------------------------------------------------------- /config/officehome/officehome_Pr-Ar-p06.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Product->Art:p06", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Product", 8 | "target": "Art", 9 | "fewshot": "p06", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 10, 29 | "load_weight_epoch": 0 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.95, 33 | "thres_tgt": 0.98, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-src", 38 | "proto-tgt", 39 | "I2C-cross", 40 | "semi-condentmax", 41 | "semi-entmin", 42 | "tgt-condentmax", 43 | "tgt-entmin" 44 | ], 45 | "weight": [1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 46 | "clus": { 47 | "kmeans_freq": 1, 48 | "type": ["each"], 49 | "n_k": 15, 50 | "k": [65, 130] 51 | } 52 | }, 53 | "confidence_params": 54 | { 55 | "confidence_ratio": 0.75, 56 | "lambda_mixed": 1.0, 57 | "mixup_alpha": 0.75, 58 | "lambda_target_mixed": 0.1, 59 | "target_inside_ratio": 0.03, 60 | "target_inside_confidence": 0.5 61 | }, 62 | "seed": 1994160032 63 | } 64 | -------------------------------------------------------------------------------- /config/officehome/officehome_Pr-Cl-p03.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Product->Clipart:p03", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Product", 8 | "target": "Clipart", 9 | "fewshot": "p03", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 30, 29 | "load_weight_epoch": 0 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.95, 33 | "thres_tgt": 0.98, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-src", 38 | "proto-tgt", 39 | "I2C-cross", 40 | "semi-condentmax", 41 | "semi-entmin", 42 | "tgt-condentmax", 43 | "tgt-entmin" 44 | ], 45 | "weight": [1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 46 | "clus": { 47 | "kmeans_freq": 1, 48 | "type": ["each"], 49 | "n_k": 15, 50 | "k": [65, 130] 51 | } 52 | }, 53 | "confidence_params": 54 | { 55 | "confidence_ratio": 0.25, 56 | "lambda_mixed": 1.0, 57 | "mixup_alpha": 0.75, 58 | "lambda_target_mixed": 0.1, 59 | "target_inside_ratio": 0.03, 60 | "target_inside_confidence": 0.75 61 | }, 62 | "seed": 788663727 63 | } 64 | -------------------------------------------------------------------------------- /config/officehome/officehome_Pr-Cl-p06.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Product->Clipart:p06", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Product", 8 | "target": "Clipart", 9 | "fewshot": "p06", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 30, 29 | "load_weight_epoch": 0 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.95, 33 | "thres_tgt": 0.98, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "proto-src", 38 | "proto-tgt", 39 | "I2C-cross", 40 | "semi-condentmax", 41 | "semi-entmin", 42 | "tgt-condentmax", 43 | "tgt-entmin" 44 | ], 45 | "weight": [1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 46 | "clus": { 47 | "kmeans_freq": 1, 48 | "type": ["each"], 49 | "n_k": 15, 50 | "k": [65, 130] 51 | } 52 | }, 53 | "confidence_params": 54 | { 55 | "confidence_ratio": 0.75, 56 | "lambda_mixed": 1.0, 57 | "mixup_alpha": 0.75, 58 | "lambda_target_mixed": 0.1, 59 | "target_inside_ratio": 0.2, 60 | "target_inside_confidence": 0.75 61 | }, 62 | "seed": 1255563890 63 | } 64 | -------------------------------------------------------------------------------- /config/officehome/officehome_Pr-Rw-p03.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Product->Real:p03", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Product", 8 | "target": "Real", 9 | "fewshot": "p03", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 30 29 | }, 30 | "loss_params": { 31 | "thres_src": 0.95, 32 | "thres_tgt": 0.99, 33 | "temp": 0.1, 34 | "loss": [ 35 | "cls-so", 36 | "proto-src", 37 | "proto-tgt", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [ 45 | 1, 46 | 1, 47 | 0.5, 48 | 1, 49 | 0.5, 50 | 0.05, 51 | 0.5, 52 | 0.05 53 | ], 54 | "clus": { 55 | "kmeans_freq": 1, 56 | "type": ["each"], 57 | "n_k": 10, 58 | "k": [65, 130, 130] 59 | } 60 | }, 61 | "confidence_params": 62 | { 63 | "confidence_ratio": 0.75, 64 | "lambda_mixed": 1.0, 65 | "mixup_alpha": 0.75, 66 | "lambda_target_mixed": 0.1, 67 | "target_inside_ratio": 0.2, 68 | "target_inside_confidence": 0.75 69 | }, 70 | "seed": 3376073 71 | } 72 | -------------------------------------------------------------------------------- /config/officehome/officehome_Pr-Rw-p06.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Product->Real:p06", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Product", 8 | "target": "Real", 9 | "fewshot": "p06", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 30 29 | }, 30 | "loss_params": { 31 | "thres_src": 0.95, 32 | "thres_tgt": 0.99, 33 | "temp": 0.1, 34 | "loss": [ 35 | "cls-so", 36 | "proto-src", 37 | "proto-tgt", 38 | "I2C-cross", 39 | "semi-condentmax", 40 | "semi-entmin", 41 | "tgt-condentmax", 42 | "tgt-entmin" 43 | ], 44 | "weight": [ 45 | 1, 46 | 1, 47 | 0.5, 48 | 1, 49 | 0.5, 50 | 0.05, 51 | 0.5, 52 | 0.05 53 | ], 54 | "clus": { 55 | "kmeans_freq": 1, 56 | "type": ["each"], 57 | "n_k": 10, 58 | "k": [65, 130, 130] 59 | } 60 | }, 61 | "confidence_params": 62 | { 63 | "confidence_ratio": 0.75, 64 | "lambda_mixed": 1.0, 65 | "mixup_alpha": 0.75, 66 | "lambda_target_mixed": 0.1, 67 | "target_inside_ratio": 0.25, 68 | "target_inside_confidence": 0.75 69 | }, 70 | "seed": 1043567647 71 | } 72 | -------------------------------------------------------------------------------- /config/officehome/officehome_Rw-Ar-p03.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Real->Art:p03", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Real", 8 | "target": "Art", 9 | "fewshot": "p03", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 10, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "cls-info", 38 | "proto-src", 39 | "proto-tgt", 40 | "I2C-cross", 41 | "semi-condentmax", 42 | "semi-entmin", 43 | "tgt-condentmax", 44 | "tgt-entmin" 45 | ], 46 | "weight": [1, 0.1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 47 | "clus": { 48 | "kmeans_freq": 1, 49 | "type": ["each"], 50 | "n_k": 15, 51 | "k": [65, 130] 52 | } 53 | }, 54 | "confidence_params": 55 | { 56 | "confidence_ratio": 0.75, 57 | "lambda_mixed": 1.0, 58 | "mixup_alpha": 0.75, 59 | "lambda_target_mixed": 0.1, 60 | "target_inside_ratio": 0.05, 61 | "target_inside_confidence": 0.75 62 | }, 63 | "seed": 1324910387 64 | } 65 | -------------------------------------------------------------------------------- /config/officehome/officehome_Rw-Ar-p06.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Real->Art:p06", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Real", 8 | "target": "Art", 9 | "fewshot": "p06", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 10, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "cls-info", 38 | "proto-src", 39 | "proto-tgt", 40 | "I2C-cross", 41 | "semi-condentmax", 42 | "semi-entmin", 43 | "tgt-condentmax", 44 | "tgt-entmin" 45 | ], 46 | "weight": [1, 0.1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 47 | "clus": { 48 | "kmeans_freq": 1, 49 | "type": ["each"], 50 | "n_k": 15, 51 | "k": [65, 130] 52 | } 53 | }, 54 | "confidence_params": 55 | { 56 | "confidence_ratio": 0.75, 57 | "lambda_mixed": 1.0, 58 | "mixup_alpha": 0.75, 59 | "lambda_target_mixed": 0.1, 60 | "target_inside_ratio": 0.05, 61 | "target_inside_confidence": 0.95 62 | }, 63 | "seed": 480703470 64 | } 65 | -------------------------------------------------------------------------------- /config/officehome/officehome_Rw-Cl-p03.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Real->Clipart:p03", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Real", 8 | "target": "Clipart", 9 | "fewshot": "p03", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 10, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "cls-info", 38 | "proto-src", 39 | "proto-tgt", 40 | "I2C-cross", 41 | "semi-condentmax", 42 | "semi-entmin", 43 | "tgt-condentmax", 44 | "tgt-entmin" 45 | ], 46 | "weight": [1, 0.1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 47 | "clus": { 48 | "kmeans_freq": 1, 49 | "type": ["each"], 50 | "n_k": 15, 51 | "k": [65, 130] 52 | } 53 | }, 54 | "confidence_params": 55 | { 56 | "confidence_ratio": 0.25, 57 | "lambda_mixed": 1.0, 58 | "mixup_alpha": 0.75, 59 | "lambda_target_mixed": 0.1, 60 | "target_inside_ratio": 0.03, 61 | "target_inside_confidence": 0.75 62 | }, 63 | "seed": 979818723 64 | } 65 | -------------------------------------------------------------------------------- /config/officehome/officehome_Rw-Cl-p06.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Real->Clipart:p06", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Real", 8 | "target": "Clipart", 9 | "fewshot": "p06", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 10, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "cls-info", 38 | "proto-src", 39 | "proto-tgt", 40 | "I2C-cross", 41 | "semi-condentmax", 42 | "semi-entmin", 43 | "tgt-condentmax", 44 | "tgt-entmin" 45 | ], 46 | "weight": [1, 0.1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 47 | "clus": { 48 | "kmeans_freq": 1, 49 | "type": ["each"], 50 | "n_k": 15, 51 | "k": [65, 130] 52 | } 53 | }, 54 | "confidence_params": 55 | { 56 | "confidence_ratio": 0.25, 57 | "lambda_mixed": 1.0, 58 | "mixup_alpha": 0.75, 59 | "lambda_target_mixed": 0.1, 60 | "target_inside_ratio": 0.2, 61 | "target_inside_confidence": 0.75 62 | }, 63 | "seed": 1337 64 | } 65 | -------------------------------------------------------------------------------- /config/officehome/officehome_Rw-Pr-p03.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Real->Product:p03", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Real", 8 | "target": "Product", 9 | "fewshot": "p03", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 10, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "cls-info", 38 | "proto-src", 39 | "proto-tgt", 40 | "I2C-cross", 41 | "semi-condentmax", 42 | "semi-entmin", 43 | "tgt-condentmax", 44 | "tgt-entmin" 45 | ], 46 | "weight": [1, 0.1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 47 | "clus": { 48 | "kmeans_freq": 1, 49 | "type": ["each"], 50 | "n_k": 15, 51 | "k": [65, 130] 52 | } 53 | }, 54 | "confidence_params": 55 | { 56 | "confidence_ratio": 0.75, 57 | "lambda_mixed": 1.0, 58 | "mixup_alpha": 0.75, 59 | "lambda_target_mixed": 0.1, 60 | "target_inside_ratio": 0.03, 61 | "target_inside_confidence": 0.95 62 | }, 63 | "seed": 1882712603 64 | } 65 | -------------------------------------------------------------------------------- /config/officehome/officehome_Rw-Pr-p06.json: -------------------------------------------------------------------------------- 1 | { 2 | "exp_base": "./exps", 3 | "exp_name": "officehome", 4 | "exp_id": "Real->Product:p06", 5 | "data_params": { 6 | "name": "office_home", 7 | "source": "Real", 8 | "target": "Product", 9 | "fewshot": "p06", 10 | "aug": "aug_0" 11 | }, 12 | "num_epochs": 500, 13 | "steps_epoch": null, 14 | "optim_params": { 15 | "learning_rate": 0.01, 16 | "conv_lr_ratio": 0.1, 17 | "patience": 4, 18 | "batch_size_lbd": 32, 19 | "batch_size": 64, 20 | "decay": true, 21 | "weight_decay": 5e-4, 22 | "cls_update": true 23 | }, 24 | "model_params": { 25 | "out_dim": 512, 26 | "version": "pretrain-resnet50", 27 | "load_weight": "src-tgt", 28 | "load_weight_thres": 10, 29 | "load_weight_epoch": 5 30 | }, 31 | "loss_params": { 32 | "thres_src": 0.99, 33 | "thres_tgt": 0.99, 34 | "temp": 0.1, 35 | "loss": [ 36 | "cls-so", 37 | "cls-info", 38 | "proto-src", 39 | "proto-tgt", 40 | "I2C-cross", 41 | "semi-condentmax", 42 | "semi-entmin", 43 | "tgt-condentmax", 44 | "tgt-entmin" 45 | ], 46 | "weight": [1, 0.1, 1, 0.5, 1, 0.5, 0.05, 0.5, 0.05], 47 | "clus": { 48 | "kmeans_freq": 1, 49 | "type": ["each"], 50 | "n_k": 15, 51 | "k": [65, 130] 52 | } 53 | }, 54 | "confidence_params": 55 | { 56 | "confidence_ratio": 0.75, 57 | "lambda_mixed": 1.0, 58 | "mixup_alpha": 0.75, 59 | "lambda_target_mixed": 0.1, 60 | "target_inside_ratio": 0.2, 61 | "target_inside_confidence": 0.95 62 | }, 63 | "seed": 1616959347 64 | } 65 | -------------------------------------------------------------------------------- /data/splits/domainnet/clipart_labeled_1.txt: -------------------------------------------------------------------------------- 1 | clipart/aircraft_carrier/clipart_001_000017.jpg 0 2 | clipart/alarm_clock/clipart_003_000058.jpg 1 3 | clipart/ant/clipart_007_000030.jpg 2 4 | clipart/anvil/clipart_008_000072.jpg 3 5 | clipart/asparagus/clipart_011_000042.jpg 4 6 | clipart/axe/clipart_012_000014.jpg 5 7 | clipart/banana/clipart_014_000022.jpg 6 8 | clipart/basket/clipart_019_000028.jpg 7 9 | clipart/bathtub/clipart_022_000093.jpg 8 10 | clipart/bear/clipart_024_000075.jpg 9 11 | clipart/bee/clipart_027_000169.jpg 10 12 | clipart/bird/clipart_032_000046.jpg 11 13 | clipart/blackberry/clipart_034_000034.jpg 12 14 | clipart/blueberry/clipart_035_000153.jpg 13 15 | clipart/bottlecap/clipart_038_000029.jpg 14 16 | clipart/broccoli/clipart_044_000069.jpg 15 17 | clipart/bus/clipart_048_000054.jpg 16 18 | clipart/butterfly/clipart_050_000074.jpg 17 19 | clipart/cactus/clipart_051_000090.jpg 18 20 | clipart/cake/clipart_052_000108.jpg 19 21 | clipart/calculator/clipart_053_000034.jpg 20 22 | clipart/camel/clipart_055_000130.jpg 21 23 | clipart/camera/clipart_056_000050.jpg 22 24 | clipart/candle/clipart_059_000102.jpg 23 25 | clipart/cannon/clipart_060_000006.jpg 24 26 | clipart/canoe/clipart_061_000019.jpg 25 27 | clipart/carrot/clipart_063_000029.jpg 26 28 | clipart/castle/clipart_064_000028.jpg 27 29 | clipart/cat/clipart_065_000011.jpg 28 30 | clipart/ceiling_fan/clipart_066_000016.jpg 29 31 | clipart/cello/clipart_067_000062.jpg 30 32 | clipart/cell_phone/clipart_068_000010.jpg 31 33 | clipart/chair/clipart_069_000091.jpg 32 34 | clipart/chandelier/clipart_070_000109.jpg 33 35 | clipart/coffee_cup/clipart_076_000203.jpg 34 36 | clipart/compass/clipart_077_000128.jpg 35 37 | clipart/computer/clipart_078_000073.jpg 36 38 | clipart/cow/clipart_082_000141.jpg 37 39 | clipart/crab/clipart_083_000002.jpg 38 40 | clipart/crocodile/clipart_085_000030.jpg 39 41 | clipart/cruise_ship/clipart_087_000022.jpg 40 42 | clipart/dog/clipart_092_000042.jpg 41 43 | clipart/dolphin/clipart_093_000014.jpg 42 44 | clipart/dragon/clipart_096_000085.jpg 43 45 | clipart/drums/clipart_099_000098.jpg 44 46 | clipart/duck/clipart_100_000089.jpg 45 47 | clipart/dumbbell/clipart_101_000318.jpg 46 48 | clipart/elephant/clipart_104_000082.jpg 47 49 | clipart/eyeglasses/clipart_108_000008.jpg 48 50 | clipart/feather/clipart_111_000154.jpg 49 51 | clipart/fence/clipart_112_000130.jpg 50 52 | clipart/fish/clipart_117_000054.jpg 51 53 | clipart/flamingo/clipart_118_000234.jpg 52 54 | clipart/flower/clipart_122_000195.jpg 53 55 | clipart/foot/clipart_124_000009.jpg 54 56 | clipart/fork/clipart_125_000126.jpg 55 57 | clipart/frog/clipart_126_000037.jpg 56 58 | clipart/giraffe/clipart_130_000014.jpg 57 59 | clipart/goatee/clipart_131_000251.jpg 58 60 | clipart/grapes/clipart_133_000080.jpg 59 61 | clipart/guitar/clipart_135_000031.jpg 60 62 | clipart/hammer/clipart_137_000147.jpg 61 63 | clipart/helicopter/clipart_143_000053.jpg 62 64 | clipart/helmet/clipart_144_000126.jpg 63 65 | clipart/horse/clipart_148_000089.jpg 64 66 | clipart/kangaroo/clipart_160_000053.jpg 65 67 | clipart/lantern/clipart_166_000022.jpg 66 68 | clipart/laptop/clipart_167_000021.jpg 67 69 | clipart/leaf/clipart_168_000004.jpg 68 70 | clipart/lion/clipart_175_000041.jpg 69 71 | clipart/lipstick/clipart_176_000043.jpg 70 72 | clipart/lobster/clipart_177_000080.jpg 71 73 | clipart/microphone/clipart_185_000085.jpg 72 74 | clipart/monkey/clipart_187_000038.jpg 73 75 | clipart/mosquito/clipart_189_000049.jpg 74 76 | clipart/mouse/clipart_192_000013.jpg 75 77 | clipart/mug/clipart_195_000151.jpg 76 78 | clipart/mushroom/clipart_196_000134.jpg 77 79 | clipart/onion/clipart_203_000087.jpg 78 80 | clipart/panda/clipart_209_000047.jpg 79 81 | clipart/peanut/clipart_215_000024.jpg 80 82 | clipart/pear/clipart_216_000052.jpg 81 83 | clipart/peas/clipart_217_000034.jpg 82 84 | clipart/pencil/clipart_218_000020.jpg 83 85 | clipart/penguin/clipart_219_000049.jpg 84 86 | clipart/pig/clipart_223_000091.jpg 85 87 | clipart/pillow/clipart_224_000048.jpg 86 88 | clipart/pineapple/clipart_225_000066.jpg 87 89 | clipart/potato/clipart_233_000043.jpg 88 90 | clipart/power_outlet/clipart_234_000001.jpg 89 91 | clipart/purse/clipart_235_000030.jpg 90 92 | clipart/rabbit/clipart_236_000037.jpg 91 93 | clipart/raccoon/clipart_237_000019.jpg 92 94 | clipart/rhinoceros/clipart_243_000061.jpg 93 95 | clipart/rifle/clipart_244_000074.jpg 94 96 | clipart/saxophone/clipart_251_000124.jpg 95 97 | clipart/screwdriver/clipart_255_000139.jpg 96 98 | clipart/sea_turtle/clipart_256_000009.jpg 97 99 | clipart/see_saw/clipart_257_000227.jpg 98 100 | clipart/sheep/clipart_259_000102.jpg 99 101 | clipart/shoe/clipart_260_000069.jpg 100 102 | clipart/skateboard/clipart_264_000165.jpg 101 103 | clipart/snake/clipart_270_000117.jpg 102 104 | clipart/speedboat/clipart_276_000178.jpg 103 105 | clipart/spider/clipart_277_000004.jpg 104 106 | clipart/squirrel/clipart_282_000197.jpg 105 107 | clipart/strawberry/clipart_291_000292.jpg 106 108 | clipart/streetlight/clipart_292_000139.jpg 107 109 | clipart/string_bean/clipart_293_000086.jpg 108 110 | clipart/submarine/clipart_294_000004.jpg 109 111 | clipart/swan/clipart_297_000312.jpg 110 112 | clipart/table/clipart_302_000219.jpg 111 113 | clipart/teapot/clipart_303_000095.jpg 112 114 | clipart/teddy-bear/clipart_304_000009.jpg 113 115 | clipart/television/clipart_306_000079.jpg 114 116 | clipart/The_Eiffel_Tower/clipart_309_000014.jpg 115 117 | clipart/The_Great_Wall_of_China/clipart_310_000050.jpg 116 118 | clipart/tiger/clipart_312_000005.jpg 117 119 | clipart/toe/clipart_314_000003.jpg 118 120 | clipart/train/clipart_322_000055.jpg 119 121 | clipart/truck/clipart_326_000010.jpg 120 122 | clipart/umbrella/clipart_329_000006.jpg 121 123 | clipart/vase/clipart_332_000053.jpg 122 124 | clipart/watermelon/clipart_335_000032.jpg 123 125 | clipart/whale/clipart_337_000242.jpg 124 126 | clipart/zebra/clipart_344_000111.jpg 125 127 | -------------------------------------------------------------------------------- /data/splits/domainnet/painting_labeled_1.txt: -------------------------------------------------------------------------------- 1 | painting/aircraft_carrier/painting_001_000113.jpg 0 2 | painting/alarm_clock/painting_003_000064.jpg 1 3 | painting/ant/painting_007_000234.jpg 2 4 | painting/anvil/painting_008_000082.jpg 3 5 | painting/asparagus/painting_011_000316.jpg 4 6 | painting/axe/painting_012_000050.jpg 5 7 | painting/banana/painting_014_000196.jpg 6 8 | painting/basket/painting_019_000115.jpg 7 9 | painting/bathtub/painting_022_000024.jpg 8 10 | painting/bear/painting_024_000140.jpg 9 11 | painting/bee/painting_027_000289.jpg 10 12 | painting/bird/painting_032_000022.jpg 11 13 | painting/blackberry/painting_034_000001.jpg 12 14 | painting/blueberry/painting_035_000029.jpg 13 15 | painting/bottlecap/painting_038_000299.jpg 14 16 | painting/broccoli/painting_044_000045.jpg 15 17 | painting/bus/painting_048_000055.jpg 16 18 | painting/butterfly/painting_050_000298.jpg 17 19 | painting/cactus/painting_051_000071.jpg 18 20 | painting/cake/painting_052_000062.jpg 19 21 | painting/calculator/painting_053_000006.jpg 20 22 | painting/camel/painting_055_000078.jpg 21 23 | painting/camera/painting_056_000010.jpg 22 24 | painting/candle/painting_059_000180.jpg 23 25 | painting/cannon/painting_060_000048.jpg 24 26 | painting/canoe/painting_061_000027.jpg 25 27 | painting/carrot/painting_063_000090.jpg 26 28 | painting/castle/painting_064_000113.jpg 27 29 | painting/cat/painting_065_000327.jpg 28 30 | painting/ceiling_fan/painting_066_000013.jpg 29 31 | painting/cello/painting_067_000149.jpg 30 32 | painting/cell_phone/painting_068_000049.jpg 31 33 | painting/chair/painting_069_000004.jpg 32 34 | painting/chandelier/painting_070_000043.jpg 33 35 | painting/coffee_cup/painting_076_000089.jpg 34 36 | painting/compass/painting_077_000030.jpg 35 37 | painting/computer/painting_078_000012.jpg 36 38 | painting/cow/painting_082_000082.jpg 37 39 | painting/crab/painting_083_000046.jpg 38 40 | painting/crocodile/painting_085_000115.jpg 39 41 | painting/cruise_ship/painting_087_000159.jpg 40 42 | painting/dog/painting_092_000702.jpg 41 43 | painting/dolphin/painting_093_000209.jpg 42 44 | painting/dragon/painting_096_000036.jpg 43 45 | painting/drums/painting_099_000102.jpg 44 46 | painting/duck/painting_100_000148.jpg 45 47 | painting/dumbbell/painting_101_000117.jpg 46 48 | painting/elephant/painting_104_000159.jpg 47 49 | painting/eyeglasses/painting_108_000067.jpg 48 50 | painting/feather/painting_111_000341.jpg 49 51 | painting/fence/painting_112_000037.jpg 50 52 | painting/fish/painting_117_000046.jpg 51 53 | painting/flamingo/painting_118_000048.jpg 52 54 | painting/flower/painting_122_000197.jpg 53 55 | painting/foot/painting_124_000027.jpg 54 56 | painting/fork/painting_125_000074.jpg 55 57 | painting/frog/painting_126_000065.jpg 56 58 | painting/giraffe/painting_130_000073.jpg 57 59 | painting/goatee/painting_131_000047.jpg 58 60 | painting/grapes/painting_133_000105.jpg 59 61 | painting/guitar/painting_135_000176.jpg 60 62 | painting/hammer/painting_137_000002.jpg 61 63 | painting/helicopter/painting_143_000025.jpg 62 64 | painting/helmet/painting_144_000018.jpg 63 65 | painting/horse/painting_148_000447.jpg 64 66 | painting/kangaroo/painting_160_000055.jpg 65 67 | painting/lantern/painting_166_000191.jpg 66 68 | painting/laptop/painting_167_000039.jpg 67 69 | painting/leaf/painting_168_000389.jpg 68 70 | painting/lion/painting_175_000172.jpg 69 71 | painting/lipstick/painting_176_000097.jpg 70 72 | painting/lobster/painting_177_000251.jpg 71 73 | painting/microphone/painting_185_000121.jpg 72 74 | painting/monkey/painting_187_000310.jpg 73 75 | painting/mosquito/painting_189_000012.jpg 74 76 | painting/mouse/painting_192_000433.jpg 75 77 | painting/mug/painting_195_000420.jpg 76 78 | painting/mushroom/painting_196_000155.jpg 77 79 | painting/onion/painting_203_000067.jpg 78 80 | painting/panda/painting_209_000028.jpg 79 81 | painting/peanut/painting_215_000065.jpg 80 82 | painting/pear/painting_216_000369.jpg 81 83 | painting/peas/painting_217_000076.jpg 82 84 | painting/pencil/painting_218_000104.jpg 83 85 | painting/penguin/painting_219_000205.jpg 84 86 | painting/pig/painting_223_000226.jpg 85 87 | painting/pillow/painting_224_000114.jpg 86 88 | painting/pineapple/painting_225_000132.jpg 87 89 | painting/potato/painting_233_000028.jpg 88 90 | painting/power_outlet/painting_234_000072.jpg 89 91 | painting/purse/painting_235_000024.jpg 90 92 | painting/rabbit/painting_236_000149.jpg 91 93 | painting/raccoon/painting_237_000128.jpg 92 94 | painting/rhinoceros/painting_243_000088.jpg 93 95 | painting/rifle/painting_244_000120.jpg 94 96 | painting/saxophone/painting_251_000031.jpg 95 97 | painting/screwdriver/painting_255_000068.jpg 96 98 | painting/sea_turtle/painting_256_000196.jpg 97 99 | painting/see_saw/painting_257_000025.jpg 98 100 | painting/sheep/painting_259_000314.jpg 99 101 | painting/shoe/painting_260_000198.jpg 100 102 | painting/skateboard/painting_264_000069.jpg 101 103 | painting/snake/painting_270_000263.jpg 102 104 | painting/speedboat/painting_276_000029.jpg 103 105 | painting/spider/painting_277_000088.jpg 104 106 | painting/squirrel/painting_282_000766.jpg 105 107 | painting/strawberry/painting_291_000073.jpg 106 108 | painting/streetlight/painting_292_000208.jpg 107 109 | painting/string_bean/painting_293_000044.jpg 108 110 | painting/submarine/painting_294_000115.jpg 109 111 | painting/swan/painting_297_000285.jpg 110 112 | painting/table/painting_302_000037.jpg 111 113 | painting/teapot/painting_303_000122.jpg 112 114 | painting/teddy-bear/painting_304_000120.jpg 113 115 | painting/television/painting_306_000049.jpg 114 116 | painting/The_Eiffel_Tower/painting_309_000115.jpg 115 117 | painting/The_Great_Wall_of_China/painting_310_000028.jpg 116 118 | painting/tiger/painting_312_000129.jpg 117 119 | painting/toe/painting_314_000001.jpg 118 120 | painting/train/painting_322_000212.jpg 119 121 | painting/truck/painting_326_000082.jpg 120 122 | painting/umbrella/painting_329_000236.jpg 121 123 | painting/vase/painting_332_000224.jpg 122 124 | painting/watermelon/painting_335_000235.jpg 123 125 | painting/whale/painting_337_000324.jpg 124 126 | painting/zebra/painting_344_000155.jpg 125 127 | -------------------------------------------------------------------------------- /data/splits/domainnet/real_labeled_1.txt: -------------------------------------------------------------------------------- 1 | real/aircraft_carrier/real_001_000200.jpg 0 2 | real/alarm_clock/real_003_000456.jpg 1 3 | real/ant/real_007_000269.jpg 2 4 | real/anvil/real_008_000016.jpg 3 5 | real/asparagus/real_011_000496.jpg 4 6 | real/axe/real_012_000239.jpg 5 7 | real/banana/real_014_000024.jpg 6 8 | real/basket/real_019_000440.jpg 7 9 | real/bathtub/real_022_000111.jpg 8 10 | real/bear/real_024_000425.jpg 9 11 | real/bee/real_027_000300.jpg 10 12 | real/bird/real_032_000316.jpg 11 13 | real/blackberry/real_034_000189.jpg 12 14 | real/blueberry/real_035_000655.jpg 13 15 | real/bottlecap/real_038_000311.jpg 14 16 | real/broccoli/real_044_000251.jpg 15 17 | real/bus/real_048_000524.jpg 16 18 | real/butterfly/real_050_000154.jpg 17 19 | real/cactus/real_051_000152.jpg 18 20 | real/cake/real_052_000185.jpg 19 21 | real/calculator/real_053_000268.jpg 20 22 | real/camel/real_055_000344.jpg 21 23 | real/camera/real_056_000252.jpg 22 24 | real/candle/real_059_000511.jpg 23 25 | real/cannon/real_060_000247.jpg 24 26 | real/canoe/real_061_000232.jpg 25 27 | real/carrot/real_063_000028.jpg 26 28 | real/castle/real_064_000589.jpg 27 29 | real/cat/real_065_000767.jpg 28 30 | real/ceiling_fan/real_066_000069.jpg 29 31 | real/cello/real_067_000044.jpg 30 32 | real/cell_phone/real_068_000134.jpg 31 33 | real/chair/real_069_000214.jpg 32 34 | real/chandelier/real_070_000044.jpg 33 35 | real/coffee_cup/real_076_000440.jpg 34 36 | real/compass/real_077_000005.jpg 35 37 | real/computer/real_078_000209.jpg 36 38 | real/cow/real_082_000463.jpg 37 39 | real/crab/real_083_000352.jpg 38 40 | real/crocodile/real_085_000297.jpg 39 41 | real/cruise_ship/real_087_000156.jpg 40 42 | real/dog/real_092_000208.jpg 41 43 | real/dolphin/real_093_000463.jpg 42 44 | real/dragon/real_096_000214.jpg 43 45 | real/drums/real_099_000092.jpg 44 46 | real/duck/real_100_000251.jpg 45 47 | real/dumbbell/real_101_000459.jpg 46 48 | real/elephant/real_104_000017.jpg 47 49 | real/eyeglasses/real_108_000272.jpg 48 50 | real/feather/real_111_000061.jpg 49 51 | real/fence/real_112_000080.jpg 50 52 | real/fish/real_117_000375.jpg 51 53 | real/flamingo/real_118_000337.jpg 52 54 | real/flower/real_122_000021.jpg 53 55 | real/foot/real_124_000300.jpg 54 56 | real/fork/real_125_000191.jpg 55 57 | real/frog/real_126_000227.jpg 56 58 | real/giraffe/real_130_000246.jpg 57 59 | real/goatee/real_131_000255.jpg 58 60 | real/grapes/real_133_000545.jpg 59 61 | real/guitar/real_135_000044.jpg 60 62 | real/hammer/real_137_000250.jpg 61 63 | real/helicopter/real_143_000480.jpg 62 64 | real/helmet/real_144_000019.jpg 63 65 | real/horse/real_148_000068.jpg 64 66 | real/kangaroo/real_160_000357.jpg 65 67 | real/lantern/real_166_000304.jpg 66 68 | real/laptop/real_167_000267.jpg 67 69 | real/leaf/real_168_000132.jpg 68 70 | real/lion/real_175_000073.jpg 69 71 | real/lipstick/real_176_000184.jpg 70 72 | real/lobster/real_177_000036.jpg 71 73 | real/microphone/real_185_000443.jpg 72 74 | real/monkey/real_187_000205.jpg 73 75 | real/mosquito/real_189_000473.jpg 74 76 | real/mouse/real_192_000131.jpg 75 77 | real/mug/real_195_000182.jpg 76 78 | real/mushroom/real_196_000452.jpg 77 79 | real/onion/real_203_000574.jpg 78 80 | real/panda/real_209_000009.jpg 79 81 | real/peanut/real_215_000281.jpg 80 82 | real/pear/real_216_000402.jpg 81 83 | real/peas/real_217_000097.jpg 82 84 | real/pencil/real_218_000153.jpg 83 85 | real/penguin/real_219_000567.jpg 84 86 | real/pig/real_223_000158.jpg 85 87 | real/pillow/real_224_000491.jpg 86 88 | real/pineapple/real_225_000300.jpg 87 89 | real/potato/real_233_000233.jpg 88 90 | real/power_outlet/real_234_000457.jpg 89 91 | real/purse/real_235_000479.jpg 90 92 | real/rabbit/real_236_000043.jpg 91 93 | real/raccoon/real_237_000646.jpg 92 94 | real/rhinoceros/real_243_000011.jpg 93 95 | real/rifle/real_244_000256.jpg 94 96 | real/saxophone/real_251_000374.jpg 95 97 | real/screwdriver/real_255_000373.jpg 96 98 | real/sea_turtle/real_256_000591.jpg 97 99 | real/see_saw/real_257_000125.jpg 98 100 | real/sheep/real_259_000462.jpg 99 101 | real/shoe/real_260_000411.jpg 100 102 | real/skateboard/real_264_000119.jpg 101 103 | real/snake/real_270_000027.jpg 102 104 | real/speedboat/real_276_000041.jpg 103 105 | real/spider/real_277_000458.jpg 104 106 | real/squirrel/real_282_000351.jpg 105 107 | real/strawberry/real_291_000148.jpg 106 108 | real/streetlight/real_292_000011.jpg 107 109 | real/string_bean/real_293_000457.jpg 108 110 | real/submarine/real_294_000383.jpg 109 111 | real/swan/real_297_000102.jpg 110 112 | real/table/real_302_000530.jpg 111 113 | real/teapot/real_303_000389.jpg 112 114 | real/teddy-bear/real_304_000117.jpg 113 115 | real/television/real_306_000200.jpg 114 116 | real/The_Eiffel_Tower/real_309_000463.jpg 115 117 | real/The_Great_Wall_of_China/real_310_000277.jpg 116 118 | real/tiger/real_312_000581.jpg 117 119 | real/toe/real_314_000278.jpg 118 120 | real/train/real_322_000036.jpg 119 121 | real/truck/real_326_000129.jpg 120 122 | real/umbrella/real_329_000080.jpg 121 123 | real/vase/real_332_000455.jpg 122 124 | real/watermelon/real_335_000145.jpg 123 125 | real/whale/real_337_000394.jpg 124 126 | real/zebra/real_344_000380.jpg 125 127 | -------------------------------------------------------------------------------- /data/splits/domainnet/sketch_labeled_1.txt: -------------------------------------------------------------------------------- 1 | sketch/aircraft_carrier/sketch_001_000028.jpg 0 2 | sketch/alarm_clock/sketch_003_000011.jpg 1 3 | sketch/ant/sketch_007_000002.jpg 2 4 | sketch/anvil/sketch_008_000087.jpg 3 5 | sketch/asparagus/sketch_011_000069.jpg 4 6 | sketch/axe/sketch_012_000028.jpg 5 7 | sketch/banana/sketch_014_000021.jpg 6 8 | sketch/basket/sketch_019_000174.jpg 7 9 | sketch/bathtub/sketch_022_000148.jpg 8 10 | sketch/bear/sketch_024_000067.jpg 9 11 | sketch/bee/sketch_027_000070.jpg 10 12 | sketch/bird/sketch_032_000120.jpg 11 13 | sketch/blackberry/sketch_034_000009.jpg 12 14 | sketch/blueberry/sketch_035_000079.jpg 13 15 | sketch/bottlecap/sketch_038_000121.jpg 14 16 | sketch/broccoli/sketch_044_000119.jpg 15 17 | sketch/bus/sketch_048_000196.jpg 16 18 | sketch/butterfly/sketch_050_000191.jpg 17 19 | sketch/cactus/sketch_051_000014.jpg 18 20 | sketch/cake/sketch_052_000024.jpg 19 21 | sketch/calculator/sketch_053_000037.jpg 20 22 | sketch/camel/sketch_055_000015.jpg 21 23 | sketch/camera/sketch_056_000081.jpg 22 24 | sketch/candle/sketch_059_000049.jpg 23 25 | sketch/cannon/sketch_060_000013.jpg 24 26 | sketch/canoe/sketch_061_000091.jpg 25 27 | sketch/carrot/sketch_063_000034.jpg 26 28 | sketch/castle/sketch_064_000013.jpg 27 29 | sketch/cat/sketch_065_000122.jpg 28 30 | sketch/ceiling_fan/sketch_066_000004.jpg 29 31 | sketch/cello/sketch_067_000028.jpg 30 32 | sketch/cell_phone/sketch_068_000011.jpg 31 33 | sketch/chair/sketch_069_000080.jpg 32 34 | sketch/chandelier/sketch_070_000017.jpg 33 35 | sketch/coffee_cup/sketch_076_000013.jpg 34 36 | sketch/compass/sketch_077_000014.jpg 35 37 | sketch/computer/sketch_078_000002.jpg 36 38 | sketch/cow/sketch_082_000009.jpg 37 39 | sketch/crab/sketch_083_000105.jpg 38 40 | sketch/crocodile/sketch_085_000043.jpg 39 41 | sketch/cruise_ship/sketch_087_000092.jpg 40 42 | sketch/dog/sketch_092_000194.jpg 41 43 | sketch/dolphin/sketch_093_000082.jpg 42 44 | sketch/dragon/sketch_096_000047.jpg 43 45 | sketch/drums/sketch_099_000045.jpg 44 46 | sketch/duck/sketch_100_000267.jpg 45 47 | sketch/dumbbell/sketch_101_000106.jpg 46 48 | sketch/elephant/sketch_104_000135.jpg 47 49 | sketch/eyeglasses/sketch_108_000119.jpg 48 50 | sketch/feather/sketch_111_000194.jpg 49 51 | sketch/fence/sketch_112_000016.jpg 50 52 | sketch/fish/sketch_117_000264.jpg 51 53 | sketch/flamingo/sketch_118_000016.jpg 52 54 | sketch/flower/sketch_122_000292.jpg 53 55 | sketch/foot/sketch_124_000088.jpg 54 56 | sketch/fork/sketch_125_000053.jpg 55 57 | sketch/frog/sketch_126_000082.jpg 56 58 | sketch/giraffe/sketch_130_000019.jpg 57 59 | sketch/goatee/sketch_131_000020.jpg 58 60 | sketch/grapes/sketch_133_000019.jpg 59 61 | sketch/guitar/sketch_135_000038.jpg 60 62 | sketch/hammer/sketch_137_000044.jpg 61 63 | sketch/helicopter/sketch_143_000063.jpg 62 64 | sketch/helmet/sketch_144_000081.jpg 63 65 | sketch/horse/sketch_148_000002.jpg 64 66 | sketch/kangaroo/sketch_160_000075.jpg 65 67 | sketch/lantern/sketch_166_000019.jpg 66 68 | sketch/laptop/sketch_167_000075.jpg 67 69 | sketch/leaf/sketch_168_000336.jpg 68 70 | sketch/lion/sketch_175_000088.jpg 69 71 | sketch/lipstick/sketch_176_000056.jpg 70 72 | sketch/lobster/sketch_177_000061.jpg 71 73 | sketch/microphone/sketch_185_000150.jpg 72 74 | sketch/monkey/sketch_187_000031.jpg 73 75 | sketch/mosquito/sketch_189_000038.jpg 74 76 | sketch/mouse/sketch_192_000121.jpg 75 77 | sketch/mug/sketch_195_000141.jpg 76 78 | sketch/mushroom/sketch_196_000107.jpg 77 79 | sketch/onion/sketch_203_000005.jpg 78 80 | sketch/panda/sketch_209_000019.jpg 79 81 | sketch/peanut/sketch_215_000044.jpg 80 82 | sketch/pear/sketch_216_000105.jpg 81 83 | sketch/peas/sketch_217_000028.jpg 82 84 | sketch/pencil/sketch_218_000025.jpg 83 85 | sketch/penguin/sketch_219_000021.jpg 84 86 | sketch/pig/sketch_223_000157.jpg 85 87 | sketch/pillow/sketch_224_000097.jpg 86 88 | sketch/pineapple/sketch_225_000031.jpg 87 89 | sketch/potato/sketch_233_000060.jpg 88 90 | sketch/power_outlet/sketch_234_000079.jpg 89 91 | sketch/purse/sketch_235_000067.jpg 90 92 | sketch/rabbit/sketch_236_000058.jpg 91 93 | sketch/raccoon/sketch_237_000293.jpg 92 94 | sketch/rhinoceros/sketch_243_000160.jpg 93 95 | sketch/rifle/sketch_244_000065.jpg 94 96 | sketch/saxophone/sketch_251_000280.jpg 95 97 | sketch/screwdriver/sketch_255_000041.jpg 96 98 | sketch/sea_turtle/sketch_256_000104.jpg 97 99 | sketch/see_saw/sketch_257_000106.jpg 98 100 | sketch/sheep/sketch_259_000424.jpg 99 101 | sketch/shoe/sketch_260_000601.jpg 100 102 | sketch/skateboard/sketch_264_000201.jpg 101 103 | sketch/snake/sketch_270_000298.jpg 102 104 | sketch/speedboat/sketch_276_000331.jpg 103 105 | sketch/spider/sketch_277_000530.jpg 104 106 | sketch/squirrel/sketch_282_000358.jpg 105 107 | sketch/strawberry/sketch_291_000020.jpg 106 108 | sketch/streetlight/sketch_292_000191.jpg 107 109 | sketch/string_bean/sketch_293_000010.jpg 108 110 | sketch/submarine/sketch_294_000112.jpg 109 111 | sketch/swan/sketch_297_000154.jpg 110 112 | sketch/table/sketch_302_000041.jpg 111 113 | sketch/teapot/sketch_303_000313.jpg 112 114 | sketch/teddy-bear/sketch_304_000016.jpg 113 115 | sketch/television/sketch_306_000064.jpg 114 116 | sketch/The_Eiffel_Tower/sketch_309_000177.jpg 115 117 | sketch/The_Great_Wall_of_China/sketch_310_000016.jpg 116 118 | sketch/tiger/sketch_312_000231.jpg 117 119 | sketch/toe/sketch_314_000034.jpg 118 120 | sketch/train/sketch_322_000229.jpg 119 121 | sketch/truck/sketch_326_000089.jpg 120 122 | sketch/umbrella/sketch_329_000143.jpg 121 123 | sketch/vase/sketch_332_000150.jpg 122 124 | sketch/watermelon/sketch_335_000103.jpg 123 125 | sketch/whale/sketch_337_000235.jpg 124 126 | sketch/zebra/sketch_344_000036.jpg 125 127 | -------------------------------------------------------------------------------- /data/splits/office/amazon_labeled_1.txt: -------------------------------------------------------------------------------- 1 | amazon/images/back_pack/frame_0057.jpg 0 2 | amazon/images/bike/frame_0015.jpg 1 3 | amazon/images/bike_helmet/frame_0001.jpg 2 4 | amazon/images/bookcase/frame_0012.jpg 3 5 | amazon/images/bottle/frame_0003.jpg 4 6 | amazon/images/calculator/frame_0086.jpg 5 7 | amazon/images/desk_chair/frame_0089.jpg 6 8 | amazon/images/desk_lamp/frame_0011.jpg 7 9 | amazon/images/desktop_computer/frame_0013.jpg 8 10 | amazon/images/file_cabinet/frame_0046.jpg 9 11 | amazon/images/headphones/frame_0031.jpg 10 12 | amazon/images/keyboard/frame_0003.jpg 11 13 | amazon/images/laptop_computer/frame_0004.jpg 12 14 | amazon/images/letter_tray/frame_0003.jpg 13 15 | amazon/images/mobile_phone/frame_0045.jpg 14 16 | amazon/images/monitor/frame_0083.jpg 15 17 | amazon/images/mouse/frame_0080.jpg 16 18 | amazon/images/mug/frame_0062.jpg 17 19 | amazon/images/paper_notebook/frame_0079.jpg 18 20 | amazon/images/pen/frame_0060.jpg 19 21 | amazon/images/phone/frame_0020.jpg 20 22 | amazon/images/printer/frame_0012.jpg 21 23 | amazon/images/projector/frame_0024.jpg 22 24 | amazon/images/punchers/frame_0092.jpg 23 25 | amazon/images/ring_binder/frame_0015.jpg 24 26 | amazon/images/ruler/frame_0002.jpg 25 27 | amazon/images/scissors/frame_0065.jpg 26 28 | amazon/images/speaker/frame_0063.jpg 27 29 | amazon/images/stapler/frame_0032.jpg 28 30 | amazon/images/tape_dispenser/frame_0009.jpg 29 31 | amazon/images/trash_can/frame_0060.jpg 30 32 | -------------------------------------------------------------------------------- /data/splits/office/amazon_labeled_3.txt: -------------------------------------------------------------------------------- 1 | amazon/images/back_pack/frame_0001.jpg 0 2 | amazon/images/back_pack/frame_0015.jpg 0 3 | amazon/images/back_pack/frame_0057.jpg 0 4 | amazon/images/bike/frame_0005.jpg 1 5 | amazon/images/bike/frame_0012.jpg 1 6 | amazon/images/bike/frame_0075.jpg 1 7 | amazon/images/bike_helmet/frame_0011.jpg 2 8 | amazon/images/bike_helmet/frame_0013.jpg 2 9 | amazon/images/bike_helmet/frame_0046.jpg 2 10 | amazon/images/bookcase/frame_0003.jpg 3 11 | amazon/images/bookcase/frame_0004.jpg 3 12 | amazon/images/bookcase/frame_0031.jpg 3 13 | amazon/images/bottle/frame_0002.jpg 4 14 | amazon/images/bottle/frame_0023.jpg 4 15 | amazon/images/bottle/frame_0031.jpg 4 16 | amazon/images/calculator/frame_0020.jpg 5 17 | amazon/images/calculator/frame_0060.jpg 5 18 | amazon/images/calculator/frame_0079.jpg 5 19 | amazon/images/desk_chair/frame_0012.jpg 6 20 | amazon/images/desk_chair/frame_0015.jpg 6 21 | amazon/images/desk_chair/frame_0024.jpg 6 22 | amazon/images/desk_lamp/frame_0002.jpg 7 23 | amazon/images/desk_lamp/frame_0063.jpg 7 24 | amazon/images/desk_lamp/frame_0065.jpg 7 25 | amazon/images/desktop_computer/frame_0009.jpg 8 26 | amazon/images/desktop_computer/frame_0032.jpg 8 27 | amazon/images/desktop_computer/frame_0086.jpg 8 28 | amazon/images/file_cabinet/frame_0009.jpg 9 29 | amazon/images/file_cabinet/frame_0060.jpg 9 30 | amazon/images/file_cabinet/frame_0070.jpg 9 31 | amazon/images/headphones/frame_0012.jpg 10 32 | amazon/images/headphones/frame_0077.jpg 10 33 | amazon/images/headphones/frame_0087.jpg 10 34 | amazon/images/keyboard/frame_0006.jpg 11 35 | amazon/images/keyboard/frame_0066.jpg 11 36 | amazon/images/keyboard/frame_0075.jpg 11 37 | amazon/images/laptop_computer/frame_0009.jpg 12 38 | amazon/images/laptop_computer/frame_0035.jpg 12 39 | amazon/images/laptop_computer/frame_0072.jpg 12 40 | amazon/images/letter_tray/frame_0039.jpg 13 41 | amazon/images/letter_tray/frame_0062.jpg 13 42 | amazon/images/letter_tray/frame_0083.jpg 13 43 | amazon/images/mobile_phone/frame_0032.jpg 14 44 | amazon/images/mobile_phone/frame_0035.jpg 14 45 | amazon/images/mobile_phone/frame_0046.jpg 14 46 | amazon/images/monitor/frame_0026.jpg 15 47 | amazon/images/monitor/frame_0059.jpg 15 48 | amazon/images/monitor/frame_0084.jpg 15 49 | amazon/images/mouse/frame_0003.jpg 16 50 | amazon/images/mouse/frame_0068.jpg 16 51 | amazon/images/mouse/frame_0086.jpg 16 52 | amazon/images/mug/frame_0001.jpg 17 53 | amazon/images/mug/frame_0010.jpg 17 54 | amazon/images/mug/frame_0066.jpg 17 55 | amazon/images/paper_notebook/frame_0019.jpg 18 56 | amazon/images/paper_notebook/frame_0043.jpg 18 57 | amazon/images/paper_notebook/frame_0091.jpg 18 58 | amazon/images/pen/frame_0049.jpg 19 59 | amazon/images/pen/frame_0061.jpg 19 60 | amazon/images/pen/frame_0085.jpg 19 61 | amazon/images/phone/frame_0007.jpg 20 62 | amazon/images/phone/frame_0010.jpg 20 63 | amazon/images/phone/frame_0022.jpg 20 64 | amazon/images/printer/frame_0021.jpg 21 65 | amazon/images/printer/frame_0064.jpg 21 66 | amazon/images/printer/frame_0076.jpg 21 67 | amazon/images/projector/frame_0010.jpg 22 68 | amazon/images/projector/frame_0028.jpg 22 69 | amazon/images/projector/frame_0072.jpg 22 70 | amazon/images/punchers/frame_0004.jpg 23 71 | amazon/images/punchers/frame_0019.jpg 23 72 | amazon/images/punchers/frame_0045.jpg 23 73 | amazon/images/ring_binder/frame_0015.jpg 24 74 | amazon/images/ring_binder/frame_0019.jpg 24 75 | amazon/images/ring_binder/frame_0068.jpg 24 76 | amazon/images/ruler/frame_0006.jpg 25 77 | amazon/images/ruler/frame_0040.jpg 25 78 | amazon/images/ruler/frame_0059.jpg 25 79 | amazon/images/scissors/frame_0036.jpg 26 80 | amazon/images/scissors/frame_0054.jpg 26 81 | amazon/images/scissors/frame_0085.jpg 26 82 | amazon/images/speaker/frame_0006.jpg 27 83 | amazon/images/speaker/frame_0023.jpg 27 84 | amazon/images/speaker/frame_0035.jpg 27 85 | amazon/images/stapler/frame_0020.jpg 28 86 | amazon/images/stapler/frame_0039.jpg 28 87 | amazon/images/stapler/frame_0060.jpg 28 88 | amazon/images/tape_dispenser/frame_0006.jpg 29 89 | amazon/images/tape_dispenser/frame_0050.jpg 29 90 | amazon/images/tape_dispenser/frame_0073.jpg 29 91 | amazon/images/trash_can/frame_0044.jpg 30 92 | amazon/images/trash_can/frame_0045.jpg 30 93 | amazon/images/trash_can/frame_0064.jpg 30 94 | -------------------------------------------------------------------------------- /data/splits/office/amazon_labeled_5.txt: -------------------------------------------------------------------------------- 1 | amazon/images/back_pack/frame_0001.jpg 0 2 | amazon/images/back_pack/frame_0012.jpg 0 3 | amazon/images/back_pack/frame_0015.jpg 0 4 | amazon/images/back_pack/frame_0057.jpg 0 5 | amazon/images/back_pack/frame_0075.jpg 0 6 | amazon/images/bike/frame_0005.jpg 1 7 | amazon/images/bike/frame_0011.jpg 1 8 | amazon/images/bike/frame_0013.jpg 1 9 | amazon/images/bike/frame_0031.jpg 1 10 | amazon/images/bike/frame_0046.jpg 1 11 | amazon/images/bike_helmet/frame_0003.jpg 2 12 | amazon/images/bike_helmet/frame_0004.jpg 2 13 | amazon/images/bike_helmet/frame_0045.jpg 2 14 | amazon/images/bike_helmet/frame_0060.jpg 2 15 | amazon/images/bike_helmet/frame_0062.jpg 2 16 | amazon/images/bookcase/frame_0002.jpg 3 17 | amazon/images/bookcase/frame_0012.jpg 3 18 | amazon/images/bookcase/frame_0015.jpg 3 19 | amazon/images/bookcase/frame_0020.jpg 3 20 | amazon/images/bookcase/frame_0024.jpg 3 21 | amazon/images/bottle/frame_0005.jpg 4 22 | amazon/images/bottle/frame_0016.jpg 4 23 | amazon/images/bottle/frame_0032.jpg 4 24 | amazon/images/bottle/frame_0033.jpg 4 25 | amazon/images/bottle/frame_0035.jpg 4 26 | amazon/images/calculator/frame_0009.jpg 5 27 | amazon/images/calculator/frame_0012.jpg 5 28 | amazon/images/calculator/frame_0060.jpg 5 29 | amazon/images/calculator/frame_0077.jpg 5 30 | amazon/images/calculator/frame_0087.jpg 5 31 | amazon/images/desk_chair/frame_0006.jpg 6 32 | amazon/images/desk_chair/frame_0035.jpg 6 33 | amazon/images/desk_chair/frame_0066.jpg 6 34 | amazon/images/desk_chair/frame_0072.jpg 6 35 | amazon/images/desk_chair/frame_0075.jpg 6 36 | amazon/images/desk_lamp/frame_0009.jpg 7 37 | amazon/images/desk_lamp/frame_0039.jpg 7 38 | amazon/images/desk_lamp/frame_0046.jpg 7 39 | amazon/images/desk_lamp/frame_0062.jpg 7 40 | amazon/images/desk_lamp/frame_0083.jpg 7 41 | amazon/images/desktop_computer/frame_0026.jpg 8 42 | amazon/images/desktop_computer/frame_0032.jpg 8 43 | amazon/images/desktop_computer/frame_0035.jpg 8 44 | amazon/images/desktop_computer/frame_0059.jpg 8 45 | amazon/images/desktop_computer/frame_0084.jpg 8 46 | amazon/images/file_cabinet/frame_0001.jpg 9 47 | amazon/images/file_cabinet/frame_0003.jpg 9 48 | amazon/images/file_cabinet/frame_0010.jpg 9 49 | amazon/images/file_cabinet/frame_0066.jpg 9 50 | amazon/images/file_cabinet/frame_0068.jpg 9 51 | amazon/images/headphones/frame_0019.jpg 10 52 | amazon/images/headphones/frame_0043.jpg 10 53 | amazon/images/headphones/frame_0061.jpg 10 54 | amazon/images/headphones/frame_0085.jpg 10 55 | amazon/images/headphones/frame_0091.jpg 10 56 | amazon/images/keyboard/frame_0007.jpg 11 57 | amazon/images/keyboard/frame_0010.jpg 11 58 | amazon/images/keyboard/frame_0022.jpg 11 59 | amazon/images/keyboard/frame_0049.jpg 11 60 | amazon/images/keyboard/frame_0076.jpg 11 61 | amazon/images/laptop_computer/frame_0010.jpg 12 62 | amazon/images/laptop_computer/frame_0021.jpg 12 63 | amazon/images/laptop_computer/frame_0028.jpg 12 64 | amazon/images/laptop_computer/frame_0064.jpg 12 65 | amazon/images/laptop_computer/frame_0072.jpg 12 66 | amazon/images/letter_tray/frame_0004.jpg 13 67 | amazon/images/letter_tray/frame_0015.jpg 13 68 | amazon/images/letter_tray/frame_0019.jpg 13 69 | amazon/images/letter_tray/frame_0045.jpg 13 70 | amazon/images/letter_tray/frame_0068.jpg 13 71 | amazon/images/mobile_phone/frame_0006.jpg 14 72 | amazon/images/mobile_phone/frame_0019.jpg 14 73 | amazon/images/mobile_phone/frame_0040.jpg 14 74 | amazon/images/mobile_phone/frame_0059.jpg 14 75 | amazon/images/mobile_phone/frame_0086.jpg 14 76 | amazon/images/monitor/frame_0006.jpg 15 77 | amazon/images/monitor/frame_0023.jpg 15 78 | amazon/images/monitor/frame_0036.jpg 15 79 | amazon/images/monitor/frame_0054.jpg 15 80 | amazon/images/monitor/frame_0085.jpg 15 81 | amazon/images/mouse/frame_0020.jpg 16 82 | amazon/images/mouse/frame_0035.jpg 16 83 | amazon/images/mouse/frame_0039.jpg 16 84 | amazon/images/mouse/frame_0060.jpg 16 85 | amazon/images/mouse/frame_0100.jpg 16 86 | amazon/images/mug/frame_0006.jpg 17 87 | amazon/images/mug/frame_0045.jpg 17 88 | amazon/images/mug/frame_0050.jpg 17 89 | amazon/images/mug/frame_0064.jpg 17 90 | amazon/images/mug/frame_0073.jpg 17 91 | amazon/images/paper_notebook/frame_0006.jpg 18 92 | amazon/images/paper_notebook/frame_0044.jpg 18 93 | amazon/images/paper_notebook/frame_0060.jpg 18 94 | amazon/images/paper_notebook/frame_0061.jpg 18 95 | amazon/images/paper_notebook/frame_0075.jpg 18 96 | amazon/images/pen/frame_0022.jpg 19 97 | amazon/images/pen/frame_0057.jpg 19 98 | amazon/images/pen/frame_0061.jpg 19 99 | amazon/images/pen/frame_0077.jpg 19 100 | amazon/images/pen/frame_0094.jpg 19 101 | amazon/images/phone/frame_0016.jpg 20 102 | amazon/images/phone/frame_0036.jpg 20 103 | amazon/images/phone/frame_0038.jpg 20 104 | amazon/images/phone/frame_0057.jpg 20 105 | amazon/images/phone/frame_0067.jpg 20 106 | amazon/images/printer/frame_0024.jpg 21 107 | amazon/images/printer/frame_0054.jpg 21 108 | amazon/images/printer/frame_0056.jpg 21 109 | amazon/images/printer/frame_0099.jpg 21 110 | amazon/images/printer/frame_0100.jpg 21 111 | amazon/images/projector/frame_0014.jpg 22 112 | amazon/images/projector/frame_0016.jpg 22 113 | amazon/images/projector/frame_0049.jpg 22 114 | amazon/images/projector/frame_0063.jpg 22 115 | amazon/images/projector/frame_0068.jpg 22 116 | amazon/images/punchers/frame_0025.jpg 23 117 | amazon/images/punchers/frame_0033.jpg 23 118 | amazon/images/punchers/frame_0045.jpg 23 119 | amazon/images/punchers/frame_0083.jpg 23 120 | amazon/images/punchers/frame_0085.jpg 23 121 | amazon/images/ring_binder/frame_0014.jpg 24 122 | amazon/images/ring_binder/frame_0021.jpg 24 123 | amazon/images/ring_binder/frame_0035.jpg 24 124 | amazon/images/ring_binder/frame_0075.jpg 24 125 | amazon/images/ring_binder/frame_0090.jpg 24 126 | amazon/images/ruler/frame_0010.jpg 25 127 | amazon/images/ruler/frame_0011.jpg 25 128 | amazon/images/ruler/frame_0020.jpg 25 129 | amazon/images/ruler/frame_0042.jpg 25 130 | amazon/images/ruler/frame_0058.jpg 25 131 | amazon/images/scissors/frame_0005.jpg 26 132 | amazon/images/scissors/frame_0007.jpg 26 133 | amazon/images/scissors/frame_0028.jpg 26 134 | amazon/images/scissors/frame_0037.jpg 26 135 | amazon/images/scissors/frame_0051.jpg 26 136 | amazon/images/speaker/frame_0012.jpg 27 137 | amazon/images/speaker/frame_0029.jpg 27 138 | amazon/images/speaker/frame_0041.jpg 27 139 | amazon/images/speaker/frame_0065.jpg 27 140 | amazon/images/speaker/frame_0090.jpg 27 141 | amazon/images/stapler/frame_0009.jpg 28 142 | amazon/images/stapler/frame_0033.jpg 28 143 | amazon/images/stapler/frame_0040.jpg 28 144 | amazon/images/stapler/frame_0059.jpg 28 145 | amazon/images/stapler/frame_0075.jpg 28 146 | amazon/images/tape_dispenser/frame_0010.jpg 29 147 | amazon/images/tape_dispenser/frame_0013.jpg 29 148 | amazon/images/tape_dispenser/frame_0026.jpg 29 149 | amazon/images/tape_dispenser/frame_0055.jpg 29 150 | amazon/images/tape_dispenser/frame_0089.jpg 29 151 | amazon/images/trash_can/frame_0025.jpg 30 152 | amazon/images/trash_can/frame_0031.jpg 30 153 | amazon/images/trash_can/frame_0047.jpg 30 154 | amazon/images/trash_can/frame_0051.jpg 30 155 | amazon/images/trash_can/frame_0059.jpg 30 156 | -------------------------------------------------------------------------------- /data/splits/office/dslr_labeled_1.txt: -------------------------------------------------------------------------------- 1 | dslr/images/back_pack/frame_0008.jpg 0 2 | dslr/images/bike/frame_0004.jpg 1 3 | dslr/images/bike_helmet/frame_0001.jpg 2 4 | dslr/images/bookcase/frame_0002.jpg 3 5 | dslr/images/bottle/frame_0002.jpg 4 6 | dslr/images/calculator/frame_0011.jpg 5 7 | dslr/images/desk_chair/frame_0012.jpg 6 8 | dslr/images/desk_lamp/frame_0002.jpg 7 9 | dslr/images/desktop_computer/frame_0002.jpg 8 10 | dslr/images/file_cabinet/frame_0013.jpg 9 11 | dslr/images/headphones/frame_0006.jpg 10 12 | dslr/images/keyboard/frame_0004.jpg 11 13 | dslr/images/laptop_computer/frame_0001.jpg 12 14 | dslr/images/letter_tray/frame_0001.jpg 13 15 | dslr/images/mobile_phone/frame_0026.jpg 14 16 | dslr/images/monitor/frame_0001.jpg 15 17 | dslr/images/mouse/frame_0006.jpg 16 18 | dslr/images/mug/frame_0008.jpg 17 19 | dslr/images/paper_notebook/frame_0010.jpg 18 20 | dslr/images/pen/frame_0008.jpg 19 21 | dslr/images/phone/frame_0003.jpg 20 22 | dslr/images/printer/frame_0002.jpg 21 23 | dslr/images/projector/frame_0006.jpg 22 24 | dslr/images/punchers/frame_0004.jpg 23 25 | dslr/images/ring_binder/frame_0001.jpg 24 26 | dslr/images/ruler/frame_0005.jpg 25 27 | dslr/images/scissors/frame_0016.jpg 26 28 | dslr/images/speaker/frame_0008.jpg 27 29 | dslr/images/stapler/frame_0003.jpg 28 30 | dslr/images/tape_dispenser/frame_0022.jpg 29 31 | dslr/images/trash_can/frame_0009.jpg 30 32 | -------------------------------------------------------------------------------- /data/splits/office/dslr_labeled_3.txt: -------------------------------------------------------------------------------- 1 | dslr/images/back_pack/frame_0001.jpg 0 2 | dslr/images/back_pack/frame_0002.jpg 0 3 | dslr/images/back_pack/frame_0008.jpg 0 4 | dslr/images/bike/frame_0002.jpg 1 5 | dslr/images/bike/frame_0003.jpg 1 6 | dslr/images/bike/frame_0019.jpg 1 7 | dslr/images/bike_helmet/frame_0003.jpg 2 8 | dslr/images/bike_helmet/frame_0022.jpg 2 9 | dslr/images/bike_helmet/frame_0023.jpg 2 10 | dslr/images/bookcase/frame_0002.jpg 3 11 | dslr/images/bookcase/frame_0004.jpg 3 12 | dslr/images/bookcase/frame_0006.jpg 3 13 | dslr/images/bottle/frame_0001.jpg 4 14 | dslr/images/bottle/frame_0013.jpg 4 15 | dslr/images/bottle/frame_0016.jpg 4 16 | dslr/images/calculator/frame_0001.jpg 5 17 | dslr/images/calculator/frame_0006.jpg 5 18 | dslr/images/calculator/frame_0010.jpg 5 19 | dslr/images/desk_chair/frame_0008.jpg 6 20 | dslr/images/desk_chair/frame_0010.jpg 6 21 | dslr/images/desk_chair/frame_0013.jpg 6 22 | dslr/images/desk_lamp/frame_0002.jpg 7 23 | dslr/images/desk_lamp/frame_0003.jpg 7 24 | dslr/images/desk_lamp/frame_0014.jpg 7 25 | dslr/images/desktop_computer/frame_0001.jpg 8 26 | dslr/images/desktop_computer/frame_0002.jpg 8 27 | dslr/images/desktop_computer/frame_0012.jpg 8 28 | dslr/images/file_cabinet/frame_0004.jpg 9 29 | dslr/images/file_cabinet/frame_0008.jpg 9 30 | dslr/images/file_cabinet/frame_0009.jpg 9 31 | dslr/images/headphones/frame_0002.jpg 10 32 | dslr/images/headphones/frame_0009.jpg 10 33 | dslr/images/headphones/frame_0011.jpg 10 34 | dslr/images/keyboard/frame_0002.jpg 11 35 | dslr/images/keyboard/frame_0008.jpg 11 36 | dslr/images/keyboard/frame_0009.jpg 11 37 | dslr/images/laptop_computer/frame_0002.jpg 12 38 | dslr/images/laptop_computer/frame_0017.jpg 12 39 | dslr/images/laptop_computer/frame_0019.jpg 12 40 | dslr/images/letter_tray/frame_0009.jpg 13 41 | dslr/images/letter_tray/frame_0014.jpg 13 42 | dslr/images/letter_tray/frame_0016.jpg 13 43 | dslr/images/mobile_phone/frame_0003.jpg 14 44 | dslr/images/mobile_phone/frame_0010.jpg 14 45 | dslr/images/mobile_phone/frame_0021.jpg 14 46 | dslr/images/monitor/frame_0009.jpg 15 47 | dslr/images/monitor/frame_0012.jpg 15 48 | dslr/images/monitor/frame_0016.jpg 15 49 | dslr/images/mouse/frame_0004.jpg 16 50 | dslr/images/mouse/frame_0011.jpg 16 51 | dslr/images/mouse/frame_0012.jpg 16 52 | dslr/images/mug/frame_0005.jpg 17 53 | dslr/images/mug/frame_0006.jpg 17 54 | dslr/images/mug/frame_0008.jpg 17 55 | dslr/images/paper_notebook/frame_0001.jpg 18 56 | dslr/images/paper_notebook/frame_0002.jpg 18 57 | dslr/images/paper_notebook/frame_0009.jpg 18 58 | dslr/images/pen/frame_0001.jpg 19 59 | dslr/images/pen/frame_0003.jpg 19 60 | dslr/images/pen/frame_0006.jpg 19 61 | dslr/images/phone/frame_0008.jpg 20 62 | dslr/images/phone/frame_0011.jpg 20 63 | dslr/images/phone/frame_0012.jpg 20 64 | dslr/images/printer/frame_0001.jpg 21 65 | dslr/images/printer/frame_0003.jpg 21 66 | dslr/images/printer/frame_0007.jpg 21 67 | dslr/images/projector/frame_0003.jpg 22 68 | dslr/images/projector/frame_0016.jpg 22 69 | dslr/images/projector/frame_0019.jpg 22 70 | dslr/images/punchers/frame_0003.jpg 23 71 | dslr/images/punchers/frame_0006.jpg 23 72 | dslr/images/punchers/frame_0007.jpg 23 73 | dslr/images/ring_binder/frame_0001.jpg 24 74 | dslr/images/ring_binder/frame_0003.jpg 24 75 | dslr/images/ring_binder/frame_0006.jpg 24 76 | dslr/images/ruler/frame_0001.jpg 25 77 | dslr/images/ruler/frame_0005.jpg 25 78 | dslr/images/ruler/frame_0007.jpg 25 79 | dslr/images/scissors/frame_0005.jpg 26 80 | dslr/images/scissors/frame_0010.jpg 26 81 | dslr/images/scissors/frame_0015.jpg 26 82 | dslr/images/speaker/frame_0002.jpg 27 83 | dslr/images/speaker/frame_0009.jpg 27 84 | dslr/images/speaker/frame_0014.jpg 27 85 | dslr/images/stapler/frame_0002.jpg 28 86 | dslr/images/stapler/frame_0006.jpg 28 87 | dslr/images/stapler/frame_0009.jpg 28 88 | dslr/images/tape_dispenser/frame_0005.jpg 29 89 | dslr/images/tape_dispenser/frame_0010.jpg 29 90 | dslr/images/tape_dispenser/frame_0015.jpg 29 91 | dslr/images/trash_can/frame_0001.jpg 30 92 | dslr/images/trash_can/frame_0007.jpg 30 93 | dslr/images/trash_can/frame_0010.jpg 30 94 | -------------------------------------------------------------------------------- /data/splits/office/dslr_labeled_5.txt: -------------------------------------------------------------------------------- 1 | dslr/images/back_pack/frame_0001.jpg 0 2 | dslr/images/back_pack/frame_0002.jpg 0 3 | dslr/images/back_pack/frame_0008.jpg 0 4 | dslr/images/back_pack/frame_0010.jpg 0 5 | dslr/images/back_pack/frame_0011.jpg 0 6 | dslr/images/bike/frame_0001.jpg 1 7 | dslr/images/bike/frame_0003.jpg 1 8 | dslr/images/bike/frame_0004.jpg 1 9 | dslr/images/bike/frame_0008.jpg 1 10 | dslr/images/bike/frame_0012.jpg 1 11 | dslr/images/bike_helmet/frame_0001.jpg 2 12 | dslr/images/bike_helmet/frame_0012.jpg 2 13 | dslr/images/bike_helmet/frame_0016.jpg 2 14 | dslr/images/bike_helmet/frame_0020.jpg 2 15 | dslr/images/bike_helmet/frame_0021.jpg 2 16 | dslr/images/bookcase/frame_0002.jpg 3 17 | dslr/images/bookcase/frame_0003.jpg 3 18 | dslr/images/bookcase/frame_0008.jpg 3 19 | dslr/images/bookcase/frame_0010.jpg 3 20 | dslr/images/bookcase/frame_0012.jpg 3 21 | dslr/images/bottle/frame_0001.jpg 4 22 | dslr/images/bottle/frame_0004.jpg 4 23 | dslr/images/bottle/frame_0008.jpg 4 24 | dslr/images/bottle/frame_0009.jpg 4 25 | dslr/images/bottle/frame_0014.jpg 4 26 | dslr/images/calculator/frame_0002.jpg 5 27 | dslr/images/calculator/frame_0004.jpg 5 28 | dslr/images/calculator/frame_0008.jpg 5 29 | dslr/images/calculator/frame_0009.jpg 5 30 | dslr/images/calculator/frame_0011.jpg 5 31 | dslr/images/desk_chair/frame_0001.jpg 6 32 | dslr/images/desk_chair/frame_0002.jpg 6 33 | dslr/images/desk_chair/frame_0009.jpg 6 34 | dslr/images/desk_chair/frame_0010.jpg 6 35 | dslr/images/desk_chair/frame_0011.jpg 6 36 | dslr/images/desk_lamp/frame_0002.jpg 7 37 | dslr/images/desk_lamp/frame_0005.jpg 7 38 | dslr/images/desk_lamp/frame_0009.jpg 7 39 | dslr/images/desk_lamp/frame_0011.jpg 7 40 | dslr/images/desk_lamp/frame_0014.jpg 7 41 | dslr/images/desktop_computer/frame_0004.jpg 8 42 | dslr/images/desktop_computer/frame_0005.jpg 8 43 | dslr/images/desktop_computer/frame_0006.jpg 8 44 | dslr/images/desktop_computer/frame_0008.jpg 8 45 | dslr/images/desktop_computer/frame_0014.jpg 8 46 | dslr/images/file_cabinet/frame_0004.jpg 9 47 | dslr/images/file_cabinet/frame_0008.jpg 9 48 | dslr/images/file_cabinet/frame_0009.jpg 9 49 | dslr/images/file_cabinet/frame_0011.jpg 9 50 | dslr/images/file_cabinet/frame_0015.jpg 9 51 | dslr/images/headphones/frame_0001.jpg 10 52 | dslr/images/headphones/frame_0002.jpg 10 53 | dslr/images/headphones/frame_0006.jpg 10 54 | dslr/images/headphones/frame_0009.jpg 10 55 | dslr/images/headphones/frame_0013.jpg 10 56 | dslr/images/keyboard/frame_0001.jpg 11 57 | dslr/images/keyboard/frame_0002.jpg 11 58 | dslr/images/keyboard/frame_0003.jpg 11 59 | dslr/images/keyboard/frame_0007.jpg 11 60 | dslr/images/keyboard/frame_0008.jpg 11 61 | dslr/images/laptop_computer/frame_0003.jpg 12 62 | dslr/images/laptop_computer/frame_0006.jpg 12 63 | dslr/images/laptop_computer/frame_0016.jpg 12 64 | dslr/images/laptop_computer/frame_0018.jpg 12 65 | dslr/images/laptop_computer/frame_0019.jpg 12 66 | dslr/images/letter_tray/frame_0001.jpg 13 67 | dslr/images/letter_tray/frame_0002.jpg 13 68 | dslr/images/letter_tray/frame_0003.jpg 13 69 | dslr/images/letter_tray/frame_0006.jpg 13 70 | dslr/images/letter_tray/frame_0007.jpg 13 71 | dslr/images/mobile_phone/frame_0005.jpg 14 72 | dslr/images/mobile_phone/frame_0010.jpg 14 73 | dslr/images/mobile_phone/frame_0017.jpg 14 74 | dslr/images/mobile_phone/frame_0022.jpg 14 75 | dslr/images/mobile_phone/frame_0031.jpg 14 76 | dslr/images/monitor/frame_0002.jpg 15 77 | dslr/images/monitor/frame_0009.jpg 15 78 | dslr/images/monitor/frame_0014.jpg 15 79 | dslr/images/monitor/frame_0015.jpg 15 80 | dslr/images/monitor/frame_0022.jpg 15 81 | dslr/images/mouse/frame_0001.jpg 16 82 | dslr/images/mouse/frame_0003.jpg 16 83 | dslr/images/mouse/frame_0005.jpg 16 84 | dslr/images/mouse/frame_0010.jpg 16 85 | dslr/images/mouse/frame_0011.jpg 16 86 | dslr/images/mug/frame_0001.jpg 17 87 | dslr/images/mug/frame_0003.jpg 17 88 | dslr/images/mug/frame_0004.jpg 17 89 | dslr/images/mug/frame_0006.jpg 17 90 | dslr/images/mug/frame_0007.jpg 17 91 | dslr/images/paper_notebook/frame_0004.jpg 18 92 | dslr/images/paper_notebook/frame_0005.jpg 18 93 | dslr/images/paper_notebook/frame_0006.jpg 18 94 | dslr/images/paper_notebook/frame_0008.jpg 18 95 | dslr/images/paper_notebook/frame_0010.jpg 18 96 | dslr/images/pen/frame_0001.jpg 19 97 | dslr/images/pen/frame_0005.jpg 19 98 | dslr/images/pen/frame_0007.jpg 19 99 | dslr/images/pen/frame_0008.jpg 19 100 | dslr/images/pen/frame_0010.jpg 19 101 | dslr/images/phone/frame_0003.jpg 20 102 | dslr/images/phone/frame_0005.jpg 20 103 | dslr/images/phone/frame_0008.jpg 20 104 | dslr/images/phone/frame_0009.jpg 20 105 | dslr/images/phone/frame_0012.jpg 20 106 | dslr/images/printer/frame_0002.jpg 21 107 | dslr/images/printer/frame_0003.jpg 21 108 | dslr/images/printer/frame_0005.jpg 21 109 | dslr/images/printer/frame_0007.jpg 21 110 | dslr/images/printer/frame_0013.jpg 21 111 | dslr/images/projector/frame_0004.jpg 22 112 | dslr/images/projector/frame_0013.jpg 22 113 | dslr/images/projector/frame_0014.jpg 22 114 | dslr/images/projector/frame_0016.jpg 22 115 | dslr/images/projector/frame_0017.jpg 22 116 | dslr/images/punchers/frame_0002.jpg 23 117 | dslr/images/punchers/frame_0007.jpg 23 118 | dslr/images/punchers/frame_0009.jpg 23 119 | dslr/images/punchers/frame_0012.jpg 23 120 | dslr/images/punchers/frame_0016.jpg 23 121 | dslr/images/ring_binder/frame_0003.jpg 24 122 | dslr/images/ring_binder/frame_0005.jpg 24 123 | dslr/images/ring_binder/frame_0006.jpg 24 124 | dslr/images/ring_binder/frame_0007.jpg 24 125 | dslr/images/ring_binder/frame_0010.jpg 24 126 | dslr/images/ruler/frame_0001.jpg 25 127 | dslr/images/ruler/frame_0002.jpg 25 128 | dslr/images/ruler/frame_0003.jpg 25 129 | dslr/images/ruler/frame_0004.jpg 25 130 | dslr/images/ruler/frame_0005.jpg 25 131 | dslr/images/scissors/frame_0001.jpg 26 132 | dslr/images/scissors/frame_0002.jpg 26 133 | dslr/images/scissors/frame_0007.jpg 26 134 | dslr/images/scissors/frame_0010.jpg 26 135 | dslr/images/scissors/frame_0013.jpg 26 136 | dslr/images/speaker/frame_0003.jpg 27 137 | dslr/images/speaker/frame_0008.jpg 27 138 | dslr/images/speaker/frame_0017.jpg 27 139 | dslr/images/speaker/frame_0023.jpg 27 140 | dslr/images/speaker/frame_0026.jpg 27 141 | dslr/images/stapler/frame_0003.jpg 28 142 | dslr/images/stapler/frame_0009.jpg 28 143 | dslr/images/stapler/frame_0010.jpg 28 144 | dslr/images/stapler/frame_0011.jpg 28 145 | dslr/images/stapler/frame_0015.jpg 28 146 | dslr/images/tape_dispenser/frame_0003.jpg 29 147 | dslr/images/tape_dispenser/frame_0004.jpg 29 148 | dslr/images/tape_dispenser/frame_0007.jpg 29 149 | dslr/images/tape_dispenser/frame_0013.jpg 29 150 | dslr/images/tape_dispenser/frame_0014.jpg 29 151 | dslr/images/trash_can/frame_0004.jpg 30 152 | dslr/images/trash_can/frame_0008.jpg 30 153 | dslr/images/trash_can/frame_0009.jpg 30 154 | dslr/images/trash_can/frame_0012.jpg 30 155 | dslr/images/trash_can/frame_0015.jpg 30 156 | -------------------------------------------------------------------------------- /data/splits/office/dslr_unlabeled_11.txt: -------------------------------------------------------------------------------- 1 | dslr/images/back_pack/frame_0003.jpg 0 2 | dslr/images/bike/frame_0004.jpg 1 3 | dslr/images/bike/frame_0005.jpg 1 4 | dslr/images/bike/frame_0006.jpg 1 5 | dslr/images/bike/frame_0007.jpg 1 6 | dslr/images/bike/frame_0008.jpg 1 7 | dslr/images/bike/frame_0009.jpg 1 8 | dslr/images/bike/frame_0010.jpg 1 9 | dslr/images/bike/frame_0011.jpg 1 10 | dslr/images/bike/frame_0013.jpg 1 11 | dslr/images/bike/frame_0019.jpg 1 12 | dslr/images/bike_helmet/frame_0002.jpg 2 13 | dslr/images/bike_helmet/frame_0004.jpg 2 14 | dslr/images/bike_helmet/frame_0005.jpg 2 15 | dslr/images/bike_helmet/frame_0006.jpg 2 16 | dslr/images/bike_helmet/frame_0007.jpg 2 17 | dslr/images/bike_helmet/frame_0011.jpg 2 18 | dslr/images/bike_helmet/frame_0012.jpg 2 19 | dslr/images/bike_helmet/frame_0013.jpg 2 20 | dslr/images/bike_helmet/frame_0014.jpg 2 21 | dslr/images/bike_helmet/frame_0019.jpg 2 22 | dslr/images/bike_helmet/frame_0021.jpg 2 23 | dslr/images/bike_helmet/frame_0022.jpg 2 24 | dslr/images/bike_helmet/frame_0024.jpg 2 25 | dslr/images/bookcase/frame_0007.jpg 3 26 | dslr/images/bottle/frame_0003.jpg 4 27 | dslr/images/bottle/frame_0004.jpg 4 28 | dslr/images/bottle/frame_0005.jpg 4 29 | dslr/images/bottle/frame_0007.jpg 4 30 | dslr/images/bottle/frame_0013.jpg 4 31 | dslr/images/calculator/frame_0006.jpg 5 32 | dslr/images/desk_chair/frame_0005.jpg 6 33 | dslr/images/desk_chair/frame_0007.jpg 6 34 | dslr/images/desk_lamp/frame_0006.jpg 7 35 | dslr/images/desk_lamp/frame_0008.jpg 7 36 | dslr/images/desk_lamp/frame_0009.jpg 7 37 | dslr/images/desktop_computer/frame_0002.jpg 8 38 | dslr/images/desktop_computer/frame_0003.jpg 8 39 | dslr/images/desktop_computer/frame_0014.jpg 8 40 | dslr/images/desktop_computer/frame_0015.jpg 8 41 | dslr/images/file_cabinet/frame_0006.jpg 9 42 | dslr/images/file_cabinet/frame_0007.jpg 9 43 | dslr/images/file_cabinet/frame_0010.jpg 9 44 | dslr/images/file_cabinet/frame_0014.jpg 9 45 | dslr/images/headphones/frame_0001.jpg 10 46 | dslr/images/headphones/frame_0005.jpg 10 47 | dslr/images/laptop_computer/frame_0001.jpg 12 48 | dslr/images/laptop_computer/frame_0004.jpg 12 49 | dslr/images/laptop_computer/frame_0006.jpg 12 50 | dslr/images/laptop_computer/frame_0009.jpg 12 51 | dslr/images/laptop_computer/frame_0012.jpg 12 52 | dslr/images/laptop_computer/frame_0014.jpg 12 53 | dslr/images/laptop_computer/frame_0015.jpg 12 54 | dslr/images/laptop_computer/frame_0016.jpg 12 55 | dslr/images/laptop_computer/frame_0018.jpg 12 56 | dslr/images/laptop_computer/frame_0020.jpg 12 57 | dslr/images/laptop_computer/frame_0021.jpg 12 58 | dslr/images/laptop_computer/frame_0023.jpg 12 59 | dslr/images/laptop_computer/frame_0024.jpg 12 60 | dslr/images/letter_tray/frame_0001.jpg 13 61 | dslr/images/letter_tray/frame_0008.jpg 13 62 | dslr/images/letter_tray/frame_0014.jpg 13 63 | dslr/images/letter_tray/frame_0015.jpg 13 64 | dslr/images/letter_tray/frame_0016.jpg 13 65 | dslr/images/mobile_phone/frame_0001.jpg 14 66 | dslr/images/mobile_phone/frame_0002.jpg 14 67 | dslr/images/mobile_phone/frame_0003.jpg 14 68 | dslr/images/mobile_phone/frame_0004.jpg 14 69 | dslr/images/mobile_phone/frame_0005.jpg 14 70 | dslr/images/mobile_phone/frame_0009.jpg 14 71 | dslr/images/mobile_phone/frame_0010.jpg 14 72 | dslr/images/mobile_phone/frame_0011.jpg 14 73 | dslr/images/mobile_phone/frame_0013.jpg 14 74 | dslr/images/mobile_phone/frame_0016.jpg 14 75 | dslr/images/mobile_phone/frame_0018.jpg 14 76 | dslr/images/mobile_phone/frame_0019.jpg 14 77 | dslr/images/mobile_phone/frame_0020.jpg 14 78 | dslr/images/mobile_phone/frame_0021.jpg 14 79 | dslr/images/mobile_phone/frame_0022.jpg 14 80 | dslr/images/mobile_phone/frame_0023.jpg 14 81 | dslr/images/mobile_phone/frame_0024.jpg 14 82 | dslr/images/mobile_phone/frame_0025.jpg 14 83 | dslr/images/mobile_phone/frame_0027.jpg 14 84 | dslr/images/mobile_phone/frame_0030.jpg 14 85 | dslr/images/monitor/frame_0001.jpg 15 86 | dslr/images/monitor/frame_0003.jpg 15 87 | dslr/images/monitor/frame_0004.jpg 15 88 | dslr/images/monitor/frame_0006.jpg 15 89 | dslr/images/monitor/frame_0011.jpg 15 90 | dslr/images/monitor/frame_0014.jpg 15 91 | dslr/images/monitor/frame_0017.jpg 15 92 | dslr/images/monitor/frame_0018.jpg 15 93 | dslr/images/monitor/frame_0020.jpg 15 94 | dslr/images/monitor/frame_0021.jpg 15 95 | dslr/images/monitor/frame_0022.jpg 15 96 | dslr/images/mouse/frame_0007.jpg 16 97 | dslr/images/phone/frame_0009.jpg 20 98 | dslr/images/phone/frame_0011.jpg 20 99 | dslr/images/printer/frame_0003.jpg 21 100 | dslr/images/printer/frame_0009.jpg 21 101 | dslr/images/printer/frame_0012.jpg 21 102 | dslr/images/printer/frame_0013.jpg 21 103 | dslr/images/projector/frame_0001.jpg 22 104 | dslr/images/projector/frame_0002.jpg 22 105 | dslr/images/projector/frame_0005.jpg 22 106 | dslr/images/projector/frame_0007.jpg 22 107 | dslr/images/projector/frame_0008.jpg 22 108 | dslr/images/projector/frame_0011.jpg 22 109 | dslr/images/projector/frame_0014.jpg 22 110 | dslr/images/projector/frame_0015.jpg 22 111 | dslr/images/projector/frame_0018.jpg 22 112 | dslr/images/projector/frame_0019.jpg 22 113 | dslr/images/projector/frame_0020.jpg 22 114 | dslr/images/projector/frame_0022.jpg 22 115 | dslr/images/punchers/frame_0003.jpg 23 116 | dslr/images/punchers/frame_0004.jpg 23 117 | dslr/images/punchers/frame_0005.jpg 23 118 | dslr/images/punchers/frame_0007.jpg 23 119 | dslr/images/punchers/frame_0009.jpg 23 120 | dslr/images/punchers/frame_0010.jpg 23 121 | dslr/images/punchers/frame_0014.jpg 23 122 | dslr/images/scissors/frame_0001.jpg 26 123 | dslr/images/scissors/frame_0003.jpg 26 124 | dslr/images/scissors/frame_0004.jpg 26 125 | dslr/images/scissors/frame_0006.jpg 26 126 | dslr/images/scissors/frame_0007.jpg 26 127 | dslr/images/scissors/frame_0014.jpg 26 128 | dslr/images/scissors/frame_0015.jpg 26 129 | dslr/images/speaker/frame_0003.jpg 27 130 | dslr/images/speaker/frame_0004.jpg 27 131 | dslr/images/speaker/frame_0006.jpg 27 132 | dslr/images/speaker/frame_0010.jpg 27 133 | dslr/images/speaker/frame_0014.jpg 27 134 | dslr/images/speaker/frame_0015.jpg 27 135 | dslr/images/speaker/frame_0016.jpg 27 136 | dslr/images/speaker/frame_0017.jpg 27 137 | dslr/images/speaker/frame_0018.jpg 27 138 | dslr/images/speaker/frame_0019.jpg 27 139 | dslr/images/speaker/frame_0021.jpg 27 140 | dslr/images/speaker/frame_0022.jpg 27 141 | dslr/images/speaker/frame_0024.jpg 27 142 | dslr/images/speaker/frame_0025.jpg 27 143 | dslr/images/speaker/frame_0026.jpg 27 144 | dslr/images/stapler/frame_0001.jpg 28 145 | dslr/images/stapler/frame_0002.jpg 28 146 | dslr/images/stapler/frame_0006.jpg 28 147 | dslr/images/stapler/frame_0007.jpg 28 148 | dslr/images/stapler/frame_0009.jpg 28 149 | dslr/images/stapler/frame_0011.jpg 28 150 | dslr/images/stapler/frame_0013.jpg 28 151 | dslr/images/stapler/frame_0015.jpg 28 152 | dslr/images/stapler/frame_0018.jpg 28 153 | dslr/images/stapler/frame_0019.jpg 28 154 | dslr/images/tape_dispenser/frame_0002.jpg 29 155 | dslr/images/tape_dispenser/frame_0004.jpg 29 156 | dslr/images/tape_dispenser/frame_0005.jpg 29 157 | dslr/images/tape_dispenser/frame_0007.jpg 29 158 | dslr/images/tape_dispenser/frame_0009.jpg 29 159 | dslr/images/tape_dispenser/frame_0010.jpg 29 160 | dslr/images/tape_dispenser/frame_0014.jpg 29 161 | dslr/images/tape_dispenser/frame_0016.jpg 29 162 | dslr/images/tape_dispenser/frame_0019.jpg 29 163 | dslr/images/tape_dispenser/frame_0020.jpg 29 164 | dslr/images/tape_dispenser/frame_0022.jpg 29 165 | dslr/images/trash_can/frame_0001.jpg 30 166 | dslr/images/trash_can/frame_0007.jpg 30 167 | dslr/images/trash_can/frame_0008.jpg 30 168 | dslr/images/trash_can/frame_0012.jpg 30 169 | -------------------------------------------------------------------------------- /data/splits/office/webcam_labeled_1.txt: -------------------------------------------------------------------------------- 1 | webcam/images/back_pack/frame_0025.jpg 0 2 | webcam/images/bike/frame_0015.jpg 1 3 | webcam/images/bike_helmet/frame_0004.jpg 2 4 | webcam/images/bookcase/frame_0001.jpg 3 5 | webcam/images/bottle/frame_0003.jpg 4 6 | webcam/images/calculator/frame_0030.jpg 5 7 | webcam/images/desk_chair/frame_0038.jpg 6 8 | webcam/images/desk_lamp/frame_0002.jpg 7 9 | webcam/images/desktop_computer/frame_0003.jpg 8 10 | webcam/images/file_cabinet/frame_0004.jpg 9 11 | webcam/images/headphones/frame_0025.jpg 10 12 | webcam/images/keyboard/frame_0012.jpg 11 13 | webcam/images/laptop_computer/frame_0008.jpg 12 14 | webcam/images/letter_tray/frame_0001.jpg 13 15 | webcam/images/mobile_phone/frame_0001.jpg 14 16 | webcam/images/monitor/frame_0002.jpg 15 17 | webcam/images/mouse/frame_0012.jpg 16 18 | webcam/images/mug/frame_0021.jpg 17 19 | webcam/images/paper_notebook/frame_0020.jpg 18 20 | webcam/images/pen/frame_0031.jpg 19 21 | webcam/images/phone/frame_0015.jpg 20 22 | webcam/images/printer/frame_0005.jpg 21 23 | webcam/images/projector/frame_0003.jpg 22 24 | webcam/images/punchers/frame_0006.jpg 23 25 | webcam/images/ring_binder/frame_0008.jpg 24 26 | webcam/images/ruler/frame_0001.jpg 25 27 | webcam/images/scissors/frame_0017.jpg 26 28 | webcam/images/speaker/frame_0016.jpg 27 29 | webcam/images/stapler/frame_0008.jpg 28 30 | webcam/images/tape_dispenser/frame_0003.jpg 29 31 | webcam/images/trash_can/frame_0018.jpg 30 32 | -------------------------------------------------------------------------------- /data/splits/office/webcam_labeled_3.txt: -------------------------------------------------------------------------------- 1 | webcam/images/back_pack/frame_0004.jpg 0 2 | webcam/images/back_pack/frame_0015.jpg 0 3 | webcam/images/back_pack/frame_0025.jpg 0 4 | webcam/images/bike/frame_0001.jpg 1 5 | webcam/images/bike/frame_0003.jpg 1 6 | webcam/images/bike/frame_0019.jpg 1 7 | webcam/images/bike_helmet/frame_0002.jpg 2 8 | webcam/images/bike_helmet/frame_0022.jpg 2 9 | webcam/images/bike_helmet/frame_0023.jpg 2 10 | webcam/images/bookcase/frame_0002.jpg 3 11 | webcam/images/bookcase/frame_0006.jpg 3 12 | webcam/images/bookcase/frame_0012.jpg 3 13 | webcam/images/bottle/frame_0001.jpg 4 14 | webcam/images/bottle/frame_0008.jpg 4 15 | webcam/images/bottle/frame_0015.jpg 4 16 | webcam/images/calculator/frame_0001.jpg 5 17 | webcam/images/calculator/frame_0012.jpg 5 18 | webcam/images/calculator/frame_0026.jpg 5 19 | webcam/images/desk_chair/frame_0030.jpg 6 20 | webcam/images/desk_chair/frame_0031.jpg 6 21 | webcam/images/desk_chair/frame_0040.jpg 6 22 | webcam/images/desk_lamp/frame_0003.jpg 7 23 | webcam/images/desk_lamp/frame_0005.jpg 7 24 | webcam/images/desk_lamp/frame_0006.jpg 7 25 | webcam/images/desktop_computer/frame_0001.jpg 8 26 | webcam/images/desktop_computer/frame_0004.jpg 8 27 | webcam/images/desktop_computer/frame_0017.jpg 8 28 | webcam/images/file_cabinet/frame_0003.jpg 9 29 | webcam/images/file_cabinet/frame_0008.jpg 9 30 | webcam/images/file_cabinet/frame_0016.jpg 9 31 | webcam/images/headphones/frame_0015.jpg 10 32 | webcam/images/headphones/frame_0018.jpg 10 33 | webcam/images/headphones/frame_0022.jpg 10 34 | webcam/images/keyboard/frame_0003.jpg 11 35 | webcam/images/keyboard/frame_0020.jpg 11 36 | webcam/images/keyboard/frame_0022.jpg 11 37 | webcam/images/laptop_computer/frame_0003.jpg 12 38 | webcam/images/laptop_computer/frame_0017.jpg 12 39 | webcam/images/laptop_computer/frame_0019.jpg 12 40 | webcam/images/letter_tray/frame_0002.jpg 13 41 | webcam/images/letter_tray/frame_0003.jpg 13 42 | webcam/images/letter_tray/frame_0009.jpg 13 43 | webcam/images/mobile_phone/frame_0010.jpg 14 44 | webcam/images/mobile_phone/frame_0016.jpg 14 45 | webcam/images/mobile_phone/frame_0021.jpg 14 46 | webcam/images/monitor/frame_0016.jpg 15 47 | webcam/images/monitor/frame_0018.jpg 15 48 | webcam/images/monitor/frame_0023.jpg 15 49 | webcam/images/mouse/frame_0007.jpg 16 50 | webcam/images/mouse/frame_0021.jpg 16 51 | webcam/images/mouse/frame_0030.jpg 16 52 | webcam/images/mug/frame_0015.jpg 17 53 | webcam/images/mug/frame_0017.jpg 17 54 | webcam/images/mug/frame_0022.jpg 17 55 | webcam/images/paper_notebook/frame_0001.jpg 18 56 | webcam/images/paper_notebook/frame_0003.jpg 18 57 | webcam/images/paper_notebook/frame_0017.jpg 18 58 | webcam/images/pen/frame_0001.jpg 19 59 | webcam/images/pen/frame_0010.jpg 19 60 | webcam/images/pen/frame_0022.jpg 19 61 | webcam/images/phone/frame_0007.jpg 20 62 | webcam/images/phone/frame_0011.jpg 20 63 | webcam/images/phone/frame_0016.jpg 20 64 | webcam/images/printer/frame_0002.jpg 21 65 | webcam/images/printer/frame_0003.jpg 21 66 | webcam/images/printer/frame_0006.jpg 21 67 | webcam/images/projector/frame_0006.jpg 22 68 | webcam/images/projector/frame_0016.jpg 22 69 | webcam/images/projector/frame_0019.jpg 22 70 | webcam/images/punchers/frame_0003.jpg 23 71 | webcam/images/punchers/frame_0007.jpg 23 72 | webcam/images/punchers/frame_0018.jpg 23 73 | webcam/images/ring_binder/frame_0002.jpg 24 74 | webcam/images/ring_binder/frame_0010.jpg 24 75 | webcam/images/ring_binder/frame_0023.jpg 24 76 | webcam/images/ruler/frame_0002.jpg 25 77 | webcam/images/ruler/frame_0003.jpg 25 78 | webcam/images/ruler/frame_0009.jpg 25 79 | webcam/images/scissors/frame_0010.jpg 26 80 | webcam/images/scissors/frame_0015.jpg 26 81 | webcam/images/scissors/frame_0022.jpg 26 82 | webcam/images/speaker/frame_0002.jpg 27 83 | webcam/images/speaker/frame_0009.jpg 27 84 | webcam/images/speaker/frame_0014.jpg 27 85 | webcam/images/stapler/frame_0002.jpg 28 86 | webcam/images/stapler/frame_0006.jpg 28 87 | webcam/images/stapler/frame_0022.jpg 28 88 | webcam/images/tape_dispenser/frame_0005.jpg 29 89 | webcam/images/tape_dispenser/frame_0009.jpg 29 90 | webcam/images/tape_dispenser/frame_0010.jpg 29 91 | webcam/images/trash_can/frame_0002.jpg 30 92 | webcam/images/trash_can/frame_0005.jpg 30 93 | webcam/images/trash_can/frame_0015.jpg 30 94 | -------------------------------------------------------------------------------- /data/splits/office/webcam_labeled_5.txt: -------------------------------------------------------------------------------- 1 | webcam/images/back_pack/frame_0001.jpg 0 2 | webcam/images/back_pack/frame_0003.jpg 0 3 | webcam/images/back_pack/frame_0004.jpg 0 4 | webcam/images/back_pack/frame_0015.jpg 0 5 | webcam/images/back_pack/frame_0025.jpg 0 6 | webcam/images/bike/frame_0002.jpg 1 7 | webcam/images/bike/frame_0003.jpg 1 8 | webcam/images/bike/frame_0004.jpg 1 9 | webcam/images/bike/frame_0012.jpg 1 10 | webcam/images/bike/frame_0019.jpg 1 11 | webcam/images/bike_helmet/frame_0001.jpg 2 12 | webcam/images/bike_helmet/frame_0008.jpg 2 13 | webcam/images/bike_helmet/frame_0012.jpg 2 14 | webcam/images/bike_helmet/frame_0021.jpg 2 15 | webcam/images/bike_helmet/frame_0026.jpg 2 16 | webcam/images/bookcase/frame_0003.jpg 3 17 | webcam/images/bookcase/frame_0008.jpg 3 18 | webcam/images/bookcase/frame_0010.jpg 3 19 | webcam/images/bookcase/frame_0011.jpg 3 20 | webcam/images/bookcase/frame_0012.jpg 3 21 | webcam/images/bottle/frame_0001.jpg 4 22 | webcam/images/bottle/frame_0002.jpg 4 23 | webcam/images/bottle/frame_0003.jpg 4 24 | webcam/images/bottle/frame_0012.jpg 4 25 | webcam/images/bottle/frame_0016.jpg 4 26 | webcam/images/calculator/frame_0008.jpg 5 27 | webcam/images/calculator/frame_0016.jpg 5 28 | webcam/images/calculator/frame_0017.jpg 5 29 | webcam/images/calculator/frame_0029.jpg 5 30 | webcam/images/calculator/frame_0031.jpg 5 31 | webcam/images/desk_chair/frame_0005.jpg 6 32 | webcam/images/desk_chair/frame_0006.jpg 6 33 | webcam/images/desk_chair/frame_0030.jpg 6 34 | webcam/images/desk_chair/frame_0035.jpg 6 35 | webcam/images/desk_chair/frame_0039.jpg 6 36 | webcam/images/desk_lamp/frame_0002.jpg 7 37 | webcam/images/desk_lamp/frame_0009.jpg 7 38 | webcam/images/desk_lamp/frame_0014.jpg 7 39 | webcam/images/desk_lamp/frame_0016.jpg 7 40 | webcam/images/desk_lamp/frame_0017.jpg 7 41 | webcam/images/desktop_computer/frame_0003.jpg 8 42 | webcam/images/desktop_computer/frame_0009.jpg 8 43 | webcam/images/desktop_computer/frame_0010.jpg 8 44 | webcam/images/desktop_computer/frame_0012.jpg 8 45 | webcam/images/desktop_computer/frame_0016.jpg 8 46 | webcam/images/file_cabinet/frame_0001.jpg 9 47 | webcam/images/file_cabinet/frame_0007.jpg 9 48 | webcam/images/file_cabinet/frame_0008.jpg 9 49 | webcam/images/file_cabinet/frame_0009.jpg 9 50 | webcam/images/file_cabinet/frame_0015.jpg 9 51 | webcam/images/headphones/frame_0001.jpg 10 52 | webcam/images/headphones/frame_0003.jpg 10 53 | webcam/images/headphones/frame_0005.jpg 10 54 | webcam/images/headphones/frame_0011.jpg 10 55 | webcam/images/headphones/frame_0023.jpg 10 56 | webcam/images/keyboard/frame_0002.jpg 11 57 | webcam/images/keyboard/frame_0006.jpg 11 58 | webcam/images/keyboard/frame_0013.jpg 11 59 | webcam/images/keyboard/frame_0016.jpg 11 60 | webcam/images/keyboard/frame_0022.jpg 11 61 | webcam/images/laptop_computer/frame_0003.jpg 12 62 | webcam/images/laptop_computer/frame_0006.jpg 12 63 | webcam/images/laptop_computer/frame_0016.jpg 12 64 | webcam/images/laptop_computer/frame_0018.jpg 12 65 | webcam/images/laptop_computer/frame_0019.jpg 12 66 | webcam/images/letter_tray/frame_0001.jpg 13 67 | webcam/images/letter_tray/frame_0005.jpg 13 68 | webcam/images/letter_tray/frame_0007.jpg 13 69 | webcam/images/letter_tray/frame_0012.jpg 13 70 | webcam/images/letter_tray/frame_0013.jpg 13 71 | webcam/images/mobile_phone/frame_0004.jpg 14 72 | webcam/images/mobile_phone/frame_0005.jpg 14 73 | webcam/images/mobile_phone/frame_0010.jpg 14 74 | webcam/images/mobile_phone/frame_0017.jpg 14 75 | webcam/images/mobile_phone/frame_0022.jpg 14 76 | webcam/images/monitor/frame_0003.jpg 15 77 | webcam/images/monitor/frame_0018.jpg 15 78 | webcam/images/monitor/frame_0027.jpg 15 79 | webcam/images/monitor/frame_0030.jpg 15 80 | webcam/images/monitor/frame_0043.jpg 15 81 | webcam/images/mouse/frame_0002.jpg 16 82 | webcam/images/mouse/frame_0005.jpg 16 83 | webcam/images/mouse/frame_0006.jpg 16 84 | webcam/images/mouse/frame_0009.jpg 16 85 | webcam/images/mouse/frame_0026.jpg 16 86 | webcam/images/mug/frame_0005.jpg 17 87 | webcam/images/mug/frame_0010.jpg 17 88 | webcam/images/mug/frame_0015.jpg 17 89 | webcam/images/mug/frame_0025.jpg 17 90 | webcam/images/mug/frame_0027.jpg 17 91 | webcam/images/paper_notebook/frame_0002.jpg 18 92 | webcam/images/paper_notebook/frame_0012.jpg 18 93 | webcam/images/paper_notebook/frame_0013.jpg 18 94 | webcam/images/paper_notebook/frame_0019.jpg 18 95 | webcam/images/paper_notebook/frame_0025.jpg 18 96 | webcam/images/pen/frame_0003.jpg 19 97 | webcam/images/pen/frame_0022.jpg 19 98 | webcam/images/pen/frame_0030.jpg 19 99 | webcam/images/pen/frame_0031.jpg 19 100 | webcam/images/pen/frame_0032.jpg 19 101 | webcam/images/phone/frame_0003.jpg 20 102 | webcam/images/phone/frame_0010.jpg 20 103 | webcam/images/phone/frame_0014.jpg 20 104 | webcam/images/phone/frame_0015.jpg 20 105 | webcam/images/phone/frame_0016.jpg 20 106 | webcam/images/printer/frame_0009.jpg 21 107 | webcam/images/printer/frame_0010.jpg 21 108 | webcam/images/printer/frame_0015.jpg 21 109 | webcam/images/printer/frame_0017.jpg 21 110 | webcam/images/printer/frame_0020.jpg 21 111 | webcam/images/projector/frame_0004.jpg 22 112 | webcam/images/projector/frame_0006.jpg 22 113 | webcam/images/projector/frame_0014.jpg 22 114 | webcam/images/projector/frame_0017.jpg 22 115 | webcam/images/projector/frame_0025.jpg 22 116 | webcam/images/punchers/frame_0004.jpg 23 117 | webcam/images/punchers/frame_0013.jpg 23 118 | webcam/images/punchers/frame_0016.jpg 23 119 | webcam/images/punchers/frame_0022.jpg 23 120 | webcam/images/punchers/frame_0027.jpg 23 121 | webcam/images/ring_binder/frame_0007.jpg 24 122 | webcam/images/ring_binder/frame_0011.jpg 24 123 | webcam/images/ring_binder/frame_0013.jpg 24 124 | webcam/images/ring_binder/frame_0017.jpg 24 125 | webcam/images/ring_binder/frame_0023.jpg 24 126 | webcam/images/ruler/frame_0003.jpg 25 127 | webcam/images/ruler/frame_0005.jpg 25 128 | webcam/images/ruler/frame_0006.jpg 25 129 | webcam/images/ruler/frame_0007.jpg 25 130 | webcam/images/ruler/frame_0010.jpg 25 131 | webcam/images/scissors/frame_0003.jpg 26 132 | webcam/images/scissors/frame_0007.jpg 26 133 | webcam/images/scissors/frame_0013.jpg 26 134 | webcam/images/scissors/frame_0015.jpg 26 135 | webcam/images/scissors/frame_0024.jpg 26 136 | webcam/images/speaker/frame_0002.jpg 27 137 | webcam/images/speaker/frame_0008.jpg 27 138 | webcam/images/speaker/frame_0010.jpg 27 139 | webcam/images/speaker/frame_0026.jpg 27 140 | webcam/images/speaker/frame_0029.jpg 27 141 | webcam/images/stapler/frame_0003.jpg 28 142 | webcam/images/stapler/frame_0010.jpg 28 143 | webcam/images/stapler/frame_0011.jpg 28 144 | webcam/images/stapler/frame_0017.jpg 28 145 | webcam/images/stapler/frame_0023.jpg 28 146 | webcam/images/tape_dispenser/frame_0003.jpg 29 147 | webcam/images/tape_dispenser/frame_0009.jpg 29 148 | webcam/images/tape_dispenser/frame_0015.jpg 29 149 | webcam/images/tape_dispenser/frame_0019.jpg 29 150 | webcam/images/tape_dispenser/frame_0023.jpg 29 151 | webcam/images/trash_can/frame_0004.jpg 30 152 | webcam/images/trash_can/frame_0007.jpg 30 153 | webcam/images/trash_can/frame_0013.jpg 30 154 | webcam/images/trash_can/frame_0014.jpg 30 155 | webcam/images/trash_can/frame_0020.jpg 30 156 | -------------------------------------------------------------------------------- /data/splits/office_home/Art_labeled_p03.txt: -------------------------------------------------------------------------------- 1 | Art/Alarm_Clock/00057.jpg 0 2 | Art/Backpack/00014.jpg 1 3 | Art/Batteries/00011.jpg 2 4 | Art/Bed/00022.jpg 3 5 | Art/Bike/00064.jpg 4 6 | Art/Bottle/00005.jpg 5 7 | Art/Bottle/00086.jpg 5 8 | Art/Bottle/00091.jpg 5 9 | Art/Bucket/00015.jpg 6 10 | Art/Calculator/00007.jpg 7 11 | Art/Calendar/00012.jpg 8 12 | Art/Candles/00003.jpg 9 13 | Art/Candles/00052.jpg 9 14 | Art/Chair/00024.jpg 10 15 | Art/Chair/00063.jpg 10 16 | Art/Clipboards/00021.jpg 11 17 | Art/Computer/00007.jpg 12 18 | Art/Couch/00016.jpg 13 19 | Art/Curtains/00031.jpg 14 20 | Art/Desk_Lamp/00020.jpg 15 21 | Art/Drill/00006.jpg 16 22 | Art/Eraser/00005.jpg 17 23 | Art/Exit_Sign/00003.jpg 18 24 | Art/Fan/00008.jpg 19 25 | Art/File_Cabinet/00004.jpg 20 26 | Art/Flipflops/00017.jpg 21 27 | Art/Flowers/00032.jpg 22 28 | Art/Flowers/00063.jpg 22 29 | Art/Folder/00019.jpg 23 30 | Art/Fork/00015.jpg 24 31 | Art/Glasses/00036.jpg 25 32 | Art/Hammer/00038.jpg 26 33 | Art/Helmet/00018.jpg 27 34 | Art/Helmet/00077.jpg 27 35 | Art/Kettle/00042.jpg 28 36 | Art/Keyboard/00003.jpg 29 37 | Art/Knives/00008.jpg 30 38 | Art/Knives/00066.jpg 30 39 | Art/Lamp_Shade/00024.jpg 31 40 | Art/Laptop/00005.jpg 32 41 | Art/Laptop/00036.jpg 32 42 | Art/Marker/00016.jpg 33 43 | Art/Monitor/00039.jpg 34 44 | Art/Mop/00032.jpg 35 45 | Art/Mouse/00016.jpg 36 46 | Art/Mug/00016.jpg 37 47 | Art/Notebook/00007.jpg 38 48 | Art/Oven/00014.jpg 39 49 | Art/Pan/00015.jpg 40 50 | Art/Paper_Clip/00017.jpg 41 51 | Art/Pen/00018.jpg 42 52 | Art/Pencil/00020.jpg 43 53 | Art/Postit_Notes/00014.jpg 44 54 | Art/Printer/00010.jpg 45 55 | Art/Push_Pin/00007.jpg 46 56 | Art/Radio/00037.jpg 47 57 | Art/Refrigerator/00042.jpg 48 58 | Art/Ruler/00003.jpg 49 59 | Art/Scissors/00004.jpg 50 60 | Art/Screwdriver/00006.jpg 51 61 | Art/Shelf/00022.jpg 52 62 | Art/Sink/00025.jpg 53 63 | Art/Sneakers/00038.jpg 54 64 | Art/Soda/00033.jpg 55 65 | Art/Speaker/00007.jpg 56 66 | Art/Spoon/00010.jpg 57 67 | Art/TV/00033.jpg 58 68 | Art/Table/00011.jpg 59 69 | Art/Telephone/00024.jpg 60 70 | Art/ToothBrush/00001.jpg 61 71 | Art/Toys/00002.jpg 62 72 | Art/Trash_Can/00007.jpg 63 73 | Art/Webcam/00015.jpg 64 -------------------------------------------------------------------------------- /data/splits/office_home/Art_labeled_p06.txt: -------------------------------------------------------------------------------- 1 | Art/Alarm_Clock/00001.jpg 0 2 | Art/Alarm_Clock/00012.jpg 0 3 | Art/Alarm_Clock/00015.jpg 0 4 | Art/Backpack/00037.jpg 1 5 | Art/Backpack/00014.jpg 1 6 | Art/Batteries/00011.jpg 2 7 | Art/Batteries/00023.jpg 2 8 | Art/Bed/00022.jpg 3 9 | Art/Bed/00007.jpg 3 10 | Art/Bike/00003.jpg 4 11 | Art/Bike/00004.jpg 4 12 | Art/Bike/00031.jpg 4 13 | Art/Bottle/00005.jpg 5 14 | Art/Bottle/00086.jpg 5 15 | Art/Bottle/00091.jpg 5 16 | Art/Bottle/00003.jpg 5 17 | Art/Bucket/00015.jpg 6 18 | Art/Bucket/00030.jpg 6 19 | Art/Calculator/00007.jpg 7 20 | Art/Calculator/00012.jpg 7 21 | Art/Calendar/00012.jpg 8 22 | Art/Calendar/00004.jpg 8 23 | Art/Candles/00003.jpg 9 24 | Art/Candles/00052.jpg 9 25 | Art/Candles/00063.jpg 9 26 | Art/Candles/00065.jpg 9 27 | Art/Chair/00024.jpg 10 28 | Art/Chair/00063.jpg 10 29 | Art/Chair/00060.jpg 10 30 | Art/Chair/00066.jpg 10 31 | Art/Clipboards/00021.jpg 11 32 | Art/Clipboards/00019.jpg 11 33 | Art/Computer/00007.jpg 12 34 | Art/Computer/00018.jpg 12 35 | Art/Computer/00036.jpg 12 36 | Art/Couch/00016.jpg 13 37 | Art/Couch/00031.jpg 13 38 | Art/Curtains/00031.jpg 14 39 | Art/Curtains/00023.jpg 14 40 | Art/Desk_Lamp/00020.jpg 15 41 | Art/Desk_Lamp/00021.jpg 15 42 | Art/Drill/00006.jpg 16 43 | Art/Drill/00015.jpg 16 44 | Art/Eraser/00005.jpg 17 45 | Art/Eraser/00017.jpg 17 46 | Art/Exit_Sign/00003.jpg 18 47 | Art/Exit_Sign/00017.jpg 18 48 | Art/Fan/00001.jpg 19 49 | Art/Fan/00008.jpg 19 50 | Art/Fan/00022.jpg 19 51 | Art/File_Cabinet/00004.jpg 20 52 | Art/Flipflops/00017.jpg 21 53 | Art/Flipflops/00043.jpg 21 54 | Art/Flipflops/00046.jpg 21 55 | Art/Flowers/00032.jpg 22 56 | Art/Flowers/00063.jpg 22 57 | Art/Flowers/00022.jpg 22 58 | Art/Flowers/00049.jpg 22 59 | Art/Folder/00019.jpg 23 60 | Art/Fork/00015.jpg 24 61 | Art/Fork/00011.jpg 24 62 | Art/Fork/00036.jpg 24 63 | Art/Glasses/00036.jpg 25 64 | Art/Glasses/00014.jpg 25 65 | Art/Hammer/00038.jpg 26 66 | Art/Hammer/00023.jpg 26 67 | Art/Helmet/00018.jpg 27 68 | Art/Helmet/00077.jpg 27 69 | Art/Helmet/00040.jpg 27 70 | Art/Helmet/00059.jpg 27 71 | Art/Kettle/00042.jpg 28 72 | Art/Kettle/00018.jpg 28 73 | Art/Kettle/00027.jpg 28 74 | Art/Keyboard/00003.jpg 29 75 | Art/Knives/00008.jpg 30 76 | Art/Knives/00066.jpg 30 77 | Art/Knives/00035.jpg 30 78 | Art/Knives/00039.jpg 30 79 | Art/Lamp_Shade/00024.jpg 31 80 | Art/Lamp_Shade/00010.jpg 31 81 | Art/Lamp_Shade/00030.jpg 31 82 | Art/Laptop/00005.jpg 32 83 | Art/Laptop/00036.jpg 32 84 | Art/Laptop/00049.jpg 32 85 | Art/Marker/00016.jpg 33 86 | Art/Monitor/00039.jpg 34 87 | Art/Monitor/00032.jpg 34 88 | Art/Monitor/00038.jpg 34 89 | Art/Mop/00032.jpg 35 90 | Art/Mop/00031.jpg 35 91 | Art/Mouse/00016.jpg 36 92 | Art/Mug/00016.jpg 37 93 | Art/Mug/00031.jpg 37 94 | Art/Mug/00039.jpg 37 95 | Art/Notebook/00015.jpg 38 96 | Art/Notebook/00007.jpg 38 97 | Art/Oven/00017.jpg 39 98 | Art/Oven/00014.jpg 39 99 | Art/Pan/00015.jpg 40 100 | Art/Pan/00015.jpg 40 101 | Art/Paper_Clip/00010.jpg 41 102 | Art/Paper_Clip/00017.jpg 41 103 | Art/Pen/00009.jpg 42 104 | Art/Pen/00018.jpg 42 105 | Art/Pencil/00020.jpg 43 106 | Art/Pencil/00025.jpg 43 107 | Art/Postit_Notes/00006.jpg 44 108 | Art/Postit_Notes/00014.jpg 44 109 | Art/Printer/00010.jpg 45 110 | Art/Push_Pin/00007.jpg 46 111 | Art/Radio/00007.jpg 47 112 | Art/Radio/00008.jpg 47 113 | Art/Radio/00037.jpg 47 114 | Art/Refrigerator/00042.jpg 48 115 | Art/Refrigerator/00032.jpg 48 116 | Art/Refrigerator/00043.jpg 48 117 | Art/Ruler/00003.jpg 49 118 | Art/Scissors/00004.jpg 50 119 | Art/Screwdriver/00006.jpg 51 120 | Art/Screwdriver/00012.jpg 51 121 | Art/Shelf/00022.jpg 52 122 | Art/Shelf/00018.jpg 52 123 | Art/Shelf/00038.jpg 52 124 | Art/Sink/00025.jpg 53 125 | Art/Sink/00021.jpg 53 126 | Art/Sneakers/00038.jpg 54 127 | Art/Sneakers/00006.jpg 54 128 | Art/Sneakers/00029.jpg 54 129 | Art/Soda/00033.jpg 55 130 | Art/Soda/00026.jpg 55 131 | Art/Speaker/00007.jpg 56 132 | Art/Spoon/00010.jpg 57 133 | Art/Spoon/00015.jpg 57 134 | Art/Spoon/00019.jpg 57 135 | Art/TV/00006.jpg 58 136 | Art/TV/00033.jpg 58 137 | Art/Table/00011.jpg 59 138 | Art/Telephone/00024.jpg 60 139 | Art/Telephone/00020.jpg 60 140 | Art/Telephone/00030.jpg 60 141 | Art/ToothBrush/00001.jpg 61 142 | Art/ToothBrush/00017.jpg 61 143 | Art/ToothBrush/00038.jpg 61 144 | Art/Toys/00002.jpg 62 145 | Art/Trash_Can/00007.jpg 63 146 | Art/Webcam/00015.jpg 64 147 | -------------------------------------------------------------------------------- /data/splits/office_home/Clipart_labeled_p03.txt: -------------------------------------------------------------------------------- 1 | Clipart/Alarm_Clock/00010.jpg 0 2 | Clipart/Alarm_Clock/00050.jpg 0 3 | Clipart/Backpack/00019.jpg 1 4 | Clipart/Backpack/00034.jpg 1 5 | Clipart/Batteries/00019.jpg 2 6 | Clipart/Batteries/00036.jpg 2 7 | Clipart/Bed/00025.jpg 3 8 | Clipart/Bed/00065.jpg 3 9 | Clipart/Bed/00088.jpg 3 10 | Clipart/Bike/00005.jpg 4 11 | Clipart/Bike/00013.jpg 4 12 | Clipart/Bike/00046.jpg 4 13 | Clipart/Bottle/00002.jpg 5 14 | Clipart/Bottle/00016.jpg 5 15 | Clipart/Bottle/00080.jpg 5 16 | Clipart/Bucket/00007.jpg 6 17 | Clipart/Bucket/00046.jpg 6 18 | Clipart/Calculator/00040.jpg 7 19 | Clipart/Calculator/00042.jpg 7 20 | Clipart/Calendar/00059.jpg 8 21 | Clipart/Calendar/00023.jpg 8 22 | Clipart/Candles/00015.jpg 9 23 | Clipart/Candles/00056.jpg 9 24 | Clipart/Candles/00067.jpg 9 25 | Clipart/Chair/00002.jpg 10 26 | Clipart/Chair/00017.jpg 10 27 | Clipart/Chair/00057.jpg 10 28 | Clipart/Clipboards/00032.jpg 11 29 | Clipart/Clipboards/00033.jpg 11 30 | Clipart/Computer/00013.jpg 12 31 | Clipart/Computer/00018.jpg 12 32 | Clipart/Computer/00049.jpg 12 33 | Clipart/Couch/00009.jpg 13 34 | Clipart/Couch/00060.jpg 13 35 | Clipart/Curtains/00025.jpg 14 36 | Clipart/Curtains/00039.jpg 14 37 | Clipart/Desk_Lamp/00003.jpg 15 38 | Clipart/Desk_Lamp/00021.jpg 15 39 | Clipart/Drill/00003.jpg 16 40 | Clipart/Eraser/00020.jpg 17 41 | Clipart/Exit_Sign/00026.jpg 18 42 | Clipart/Fan/00005.jpg 19 43 | Clipart/Fan/00047.jpg 19 44 | Clipart/File_Cabinet/00026.jpg 20 45 | Clipart/Flipflops/00030.jpg 21 46 | Clipart/Flowers/00032.jpg 22 47 | Clipart/Flowers/00035.jpg 22 48 | Clipart/Flowers/00045.jpg 22 49 | Clipart/Folder/00018.jpg 23 50 | Clipart/Folder/00040.jpg 23 51 | Clipart/Folder/00062.jpg 23 52 | Clipart/Fork/00034.jpg 24 53 | Clipart/Fork/00047.jpg 24 54 | Clipart/Glasses/00010.jpg 25 55 | Clipart/Glasses/00035.jpg 25 56 | Clipart/Hammer/00001.jpg 26 57 | Clipart/Hammer/00010.jpg 26 58 | Clipart/Hammer/00066.jpg 26 59 | Clipart/Helmet/00014.jpg 27 60 | Clipart/Helmet/00047.jpg 27 61 | Clipart/Kettle/00004.jpg 28 62 | Clipart/Kettle/00019.jpg 28 63 | Clipart/Keyboard/00022.jpg 29 64 | Clipart/Keyboard/00053.jpg 29 65 | Clipart/Knives/00004.jpg 30 66 | Clipart/Knives/00024.jpg 30 67 | Clipart/Lamp_Shade/00030.jpg 31 68 | Clipart/Laptop/00010.jpg 32 69 | Clipart/Laptop/00021.jpg 32 70 | Clipart/Marker/00016.jpg 33 71 | Clipart/Marker/00044.jpg 33 72 | Clipart/Monitor/00004.jpg 34 73 | Clipart/Monitor/00048.jpg 34 74 | Clipart/Monitor/00098.jpg 34 75 | Clipart/Mop/00022.jpg 35 76 | Clipart/Mouse/00015.jpg 36 77 | Clipart/Mouse/00066.jpg 36 78 | Clipart/Mug/00046.jpg 37 79 | Clipart/Mug/00059.jpg 37 80 | Clipart/Mug/00086.jpg 37 81 | Clipart/Notebook/00006.jpg 38 82 | Clipart/Notebook/00053.jpg 38 83 | Clipart/Oven/00024.jpg 39 84 | Clipart/Pan/00006.jpg 40 85 | Clipart/Pan/00043.jpg 40 86 | Clipart/Paper_Clip/00036.jpg 41 87 | Clipart/Pen/00020.jpg 42 88 | Clipart/Pen/00035.jpg 42 89 | Clipart/Pen/00039.jpg 42 90 | Clipart/Pencil/00006.jpg 43 91 | Clipart/Pencil/00020.jpg 43 92 | Clipart/Pencil/00060.jpg 43 93 | Clipart/Postit_Notes/00010.jpg 44 94 | Clipart/Printer/00026.jpg 45 95 | Clipart/Printer/00050.jpg 45 96 | Clipart/Printer/00073.jpg 45 97 | Clipart/Push_Pin/00034.jpg 46 98 | Clipart/Radio/00004.jpg 47 99 | Clipart/Refrigerator/00014.jpg 48 100 | Clipart/Ruler/00012.jpg 49 101 | Clipart/Ruler/00048.jpg 49 102 | Clipart/Scissors/00022.jpg 50 103 | Clipart/Scissors/00062.jpg 50 104 | Clipart/Scissors/00081.jpg 50 105 | Clipart/Screwdriver/00023.jpg 51 106 | Clipart/Screwdriver/00012.jpg 51 107 | Clipart/Shelf/00018.jpg 52 108 | Clipart/Sink/00014.jpg 53 109 | Clipart/Sneakers/00016.jpg 54 110 | Clipart/Soda/00008.jpg 55 111 | Clipart/Soda/00016.jpg 55 112 | Clipart/Speaker/00024.jpg 56 113 | Clipart/Speaker/00047.jpg 56 114 | Clipart/Speaker/00086.jpg 56 115 | Clipart/Spoon/00001.jpg 57 116 | Clipart/Spoon/00023.jpg 57 117 | Clipart/TV/00010.jpg 58 118 | Clipart/TV/00053.jpg 58 119 | Clipart/Table/00033.jpg 59 120 | Clipart/Table/00060.jpg 59 121 | Clipart/Telephone/00022.jpg 60 122 | Clipart/Telephone/00049.jpg 60 123 | Clipart/Telephone/00098.jpg 60 124 | Clipart/ToothBrush/00033.jpg 61 125 | Clipart/Toys/00021.jpg 62 126 | Clipart/Toys/00073.jpg 62 127 | Clipart/Toys/00087.jpg 62 128 | Clipart/Trash_Can/00006.jpg 63 129 | Clipart/Trash_Can/00018.jpg 63 130 | Clipart/Webcam/00014.jpg 64 131 | Clipart/Webcam/00039.jpg 64 132 | -------------------------------------------------------------------------------- /data/splits/office_home/Product_labeled_p03.txt: -------------------------------------------------------------------------------- 1 | Product/Alarm_Clock/00015.jpg 0 2 | Product/Alarm_Clock/00057.jpg 0 3 | Product/Backpack/00001.jpg 1 4 | Product/Backpack/00012.jpg 1 5 | Product/Batteries/00003.jpg 2 6 | Product/Batteries/00043.jpg 2 7 | Product/Bed/00005.jpg 3 8 | Product/Bed/00033.jpg 3 9 | Product/Bike/00016.jpg 4 10 | Product/Bike/00023.jpg 4 11 | Product/Bottle/00010.jpg 5 12 | Product/Bottle/00036.jpg 5 13 | Product/Bucket/00002.jpg 6 14 | Product/Bucket/00023.jpg 6 15 | Product/Calculator/00062.jpg 7 16 | Product/Calculator/00080.jpg 7 17 | Product/Calendar/00060.jpg 8 18 | Product/Calendar/00079.jpg 8 19 | Product/Candles/00006.jpg 9 20 | Product/Candles/00010.jpg 9 21 | Product/Chair/00024.jpg 10 22 | Product/Chair/00053.jpg 10 23 | Product/Chair/00092.jpg 10 24 | Product/Clipboards/00019.jpg 11 25 | Product/Clipboards/00065.jpg 11 26 | Product/Computer/00009.jpg 12 27 | Product/Computer/00032.jpg 12 28 | Product/Computer/00063.jpg 12 29 | Product/Couch/00060.jpg 13 30 | Product/Couch/00070.jpg 13 31 | Product/Couch/00086.jpg 13 32 | Product/Curtains/00009.jpg 14 33 | Product/Curtains/00012.jpg 14 34 | Product/Desk_Lamp/00066.jpg 15 35 | Product/Desk_Lamp/00075.jpg 15 36 | Product/Drill/00006.jpg 16 37 | Product/Drill/00035.jpg 16 38 | Product/Eraser/00008.jpg 17 39 | Product/Eraser/00033.jpg 17 40 | Product/Exit_Sign/00039.jpg 18 41 | Product/Exit_Sign/00062.jpg 18 42 | Product/Fan/00023.jpg 19 43 | Product/Fan/00055.jpg 19 44 | Product/File_Cabinet/00032.jpg 20 45 | Product/File_Cabinet/00035.jpg 20 46 | Product/Flipflops/00034.jpg 21 47 | Product/Flipflops/00087.jpg 21 48 | Product/Flowers/00003.jpg 22 49 | Product/Flowers/00068.jpg 22 50 | Product/Flowers/00086.jpg 22 51 | Product/Folder/00001.jpg 23 52 | Product/Folder/00010.jpg 23 53 | Product/Folder/00066.jpg 23 54 | Product/Fork/00010.jpg 24 55 | Product/Fork/00022.jpg 24 56 | Product/Glasses/00023.jpg 25 57 | Product/Glasses/00055.jpg 25 58 | Product/Hammer/00004.jpg 26 59 | Product/Hammer/00011.jpg 26 60 | Product/Helmet/00012.jpg 27 61 | Product/Helmet/00051.jpg 27 62 | Product/Helmet/00064.jpg 27 63 | Product/Kettle/00010.jpg 28 64 | Product/Kettle/00021.jpg 28 65 | Product/Keyboard/00019.jpg 29 66 | Product/Keyboard/00028.jpg 29 67 | Product/Keyboard/00072.jpg 29 68 | Product/Knives/00002.jpg 30 69 | Product/Knives/00023.jpg 30 70 | Product/Lamp_Shade/00008.jpg 31 71 | Product/Lamp_Shade/00050.jpg 31 72 | Product/Laptop/00019.jpg 32 73 | Product/Laptop/00068.jpg 32 74 | Product/Laptop/00086.jpg 32 75 | Product/Marker/00020.jpg 33 76 | Product/Marker/00030.jpg 33 77 | Product/Monitor/00006.jpg 34 78 | Product/Monitor/00036.jpg 34 79 | Product/Monitor/00054.jpg 34 80 | Product/Mop/00006.jpg 35 81 | Product/Mop/00023.jpg 35 82 | Product/Mouse/00020.jpg 36 83 | Product/Mouse/00039.jpg 36 84 | Product/Mouse/00090.jpg 36 85 | Product/Mug/00010.jpg 37 86 | Product/Notebook/00006.jpg 38 87 | Product/Notebook/00060.jpg 38 88 | Product/Notebook/00073.jpg 38 89 | Product/Oven/00045.jpg 39 90 | Product/Oven/00050.jpg 39 91 | Product/Pan/00044.jpg 40 92 | Product/Pan/00064.jpg 40 93 | Product/Paper_Clip/00038.jpg 41 94 | Product/Pen/00030.jpg 42 95 | Product/Pen/00053.jpg 42 96 | Product/Pencil/00031.jpg 43 97 | Product/Postit_Notes/00003.jpg 44 98 | Product/Printer/00022.jpg 45 99 | Product/Printer/00061.jpg 45 100 | Product/Printer/00077.jpg 45 101 | Product/Push_Pin/00029.jpg 46 102 | Product/Radio/00022.jpg 47 103 | Product/Refrigerator/00010.jpg 48 104 | Product/Refrigerator/00053.jpg 48 105 | Product/Ruler/00018.jpg 49 106 | Product/Ruler/00019.jpg 49 107 | Product/Scissors/00016.jpg 50 108 | Product/Scissors/00024.jpg 50 109 | Product/Scissors/00054.jpg 50 110 | Product/Screwdriver/00028.jpg 51 111 | Product/Shelf/00012.jpg 52 112 | Product/Sink/00025.jpg 53 113 | Product/Sneakers/00014.jpg 54 114 | Product/Sneakers/00049.jpg 54 115 | Product/Sneakers/00063.jpg 54 116 | Product/Soda/00043.jpg 55 117 | Product/Speaker/00025.jpg 56 118 | Product/Speaker/00033.jpg 56 119 | Product/Speaker/00083.jpg 56 120 | Product/Spoon/00023.jpg 57 121 | Product/TV/00014.jpg 58 122 | Product/TV/00021.jpg 58 123 | Product/Table/00005.jpg 59 124 | Product/Table/00040.jpg 59 125 | Product/Telephone/00010.jpg 60 126 | Product/Telephone/00018.jpg 60 127 | Product/ToothBrush/00010.jpg 61 128 | Product/Toys/00038.jpg 62 129 | Product/Trash_Can/00010.jpg 63 130 | Product/Trash_Can/00011.jpg 63 131 | Product/Trash_Can/00093.jpg 63 132 | Product/Webcam/00061.jpg 64 133 | Product/Webcam/00084.jpg 64 134 | -------------------------------------------------------------------------------- /data/splits/office_home/RealWorld_labeled_p03.txt: -------------------------------------------------------------------------------- 1 | RealWorld/Alarm_Clock/00001.jpg 0 2 | RealWorld/Alarm_Clock/00015.jpg 0 3 | RealWorld/Alarm_Clock/00057.jpg 0 4 | RealWorld/Backpack/00005.jpg 1 5 | RealWorld/Backpack/00012.jpg 1 6 | RealWorld/Batteries/00011.jpg 2 7 | RealWorld/Batteries/00013.jpg 2 8 | RealWorld/Bed/00041.jpg 3 9 | RealWorld/Bed/00054.jpg 3 10 | RealWorld/Bike/00003.jpg 4 11 | RealWorld/Bike/00004.jpg 4 12 | RealWorld/Bike/00045.jpg 4 13 | RealWorld/Bottle/00016.jpg 5 14 | RealWorld/Bottle/00062.jpg 5 15 | RealWorld/Bucket/00012.jpg 6 16 | RealWorld/Bucket/00020.jpg 6 17 | RealWorld/Calculator/00015.jpg 7 18 | RealWorld/Calculator/00024.jpg 7 19 | RealWorld/Calendar/00002.jpg 8 20 | RealWorld/Calendar/00065.jpg 8 21 | RealWorld/Candles/00009.jpg 9 22 | RealWorld/Candles/00032.jpg 9 23 | RealWorld/Candles/00063.jpg 9 24 | RealWorld/Chair/00060.jpg 10 25 | RealWorld/Chair/00084.jpg 10 26 | RealWorld/Chair/00086.jpg 10 27 | RealWorld/Clipboards/00009.jpg 11 28 | RealWorld/Clipboards/00024.jpg 11 29 | RealWorld/Computer/00006.jpg 12 30 | RealWorld/Computer/00035.jpg 12 31 | RealWorld/Couch/00003.jpg 13 32 | RealWorld/Couch/00059.jpg 13 33 | RealWorld/Curtains/00039.jpg 14 34 | RealWorld/Curtains/00062.jpg 14 35 | RealWorld/Desk_Lamp/00023.jpg 15 36 | RealWorld/Desk_Lamp/00055.jpg 15 37 | RealWorld/Drill/00016.jpg 16 38 | RealWorld/Drill/00018.jpg 16 39 | RealWorld/Eraser/00014.jpg 17 40 | RealWorld/Eraser/00035.jpg 17 41 | RealWorld/Exit_Sign/00059.jpg 18 42 | RealWorld/Exit_Sign/00068.jpg 18 43 | RealWorld/Fan/00002.jpg 19 44 | RealWorld/Fan/00043.jpg 19 45 | RealWorld/File_Cabinet/00011.jpg 20 46 | RealWorld/File_Cabinet/00033.jpg 20 47 | RealWorld/Flipflops/00019.jpg 21 48 | RealWorld/Flipflops/00043.jpg 21 49 | RealWorld/Flowers/00049.jpg 22 50 | RealWorld/Flowers/00061.jpg 22 51 | RealWorld/Folder/00004.jpg 23 52 | RealWorld/Folder/00021.jpg 23 53 | RealWorld/Fork/00010.jpg 24 54 | RealWorld/Glasses/00032.jpg 25 55 | RealWorld/Glasses/00038.jpg 25 56 | RealWorld/Hammer/00005.jpg 26 57 | RealWorld/Hammer/00011.jpg 26 58 | RealWorld/Helmet/00014.jpg 27 59 | RealWorld/Helmet/00036.jpg 27 60 | RealWorld/Kettle/00004.jpg 28 61 | RealWorld/Kettle/00019.jpg 28 62 | RealWorld/Keyboard/00015.jpg 29 63 | RealWorld/Keyboard/00045.jpg 29 64 | RealWorld/Knives/00019.jpg 30 65 | RealWorld/Knives/00068.jpg 30 66 | RealWorld/Lamp_Shade/00040.jpg 31 67 | RealWorld/Lamp_Shade/00059.jpg 31 68 | RealWorld/Laptop/00006.jpg 32 69 | RealWorld/Laptop/00054.jpg 32 70 | RealWorld/Marker/00008.jpg 33 71 | RealWorld/Monitor/00006.jpg 34 72 | RealWorld/Monitor/00023.jpg 34 73 | RealWorld/Mop/00008.jpg 35 74 | RealWorld/Mouse/00010.jpg 36 75 | RealWorld/Mouse/00020.jpg 36 76 | RealWorld/Mug/00010.jpg 37 77 | RealWorld/Mug/00055.jpg 37 78 | RealWorld/Notebook/00006.jpg 38 79 | RealWorld/Notebook/00060.jpg 38 80 | RealWorld/Oven/00014.jpg 39 81 | RealWorld/Oven/00042.jpg 39 82 | RealWorld/Pan/00016.jpg 40 83 | RealWorld/Paper_Clip/00044.jpg 41 84 | RealWorld/Paper_Clip/00060.jpg 41 85 | RealWorld/Pen/00006.jpg 42 86 | RealWorld/Pen/00061.jpg 42 87 | RealWorld/Pencil/00031.jpg 43 88 | RealWorld/Pencil/00051.jpg 43 89 | RealWorld/Postit_Notes/00010.jpg 44 90 | RealWorld/Postit_Notes/00041.jpg 44 91 | RealWorld/Printer/00047.jpg 45 92 | RealWorld/Printer/00050.jpg 45 93 | RealWorld/Push_Pin/00034.jpg 46 94 | RealWorld/Push_Pin/00053.jpg 46 95 | RealWorld/Radio/00038.jpg 47 96 | RealWorld/Radio/00045.jpg 47 97 | RealWorld/Refrigerator/00019.jpg 48 98 | RealWorld/Refrigerator/00041.jpg 48 99 | RealWorld/Ruler/00012.jpg 49 100 | RealWorld/Scissors/00054.jpg 50 101 | RealWorld/Scissors/00056.jpg 50 102 | RealWorld/Screwdriver/00008.jpg 51 103 | RealWorld/Screwdriver/00050.jpg 51 104 | RealWorld/Shelf/00014.jpg 52 105 | RealWorld/Shelf/00063.jpg 52 106 | RealWorld/Sink/00013.jpg 53 107 | RealWorld/Sink/00064.jpg 53 108 | RealWorld/Sneakers/00025.jpg 54 109 | RealWorld/Sneakers/00045.jpg 54 110 | RealWorld/Sneakers/00083.jpg 54 111 | RealWorld/Soda/00007.jpg 55 112 | RealWorld/Soda/00045.jpg 55 113 | RealWorld/Speaker/00021.jpg 56 114 | RealWorld/Speaker/00075.jpg 56 115 | RealWorld/Spoon/00010.jpg 57 116 | RealWorld/Spoon/00018.jpg 57 117 | RealWorld/TV/00021.jpg 58 118 | RealWorld/TV/00047.jpg 58 119 | RealWorld/Table/00018.jpg 59 120 | RealWorld/Table/00054.jpg 59 121 | RealWorld/Telephone/00010.jpg 60 122 | RealWorld/Telephone/00011.jpg 60 123 | RealWorld/ToothBrush/00011.jpg 61 124 | RealWorld/ToothBrush/00030.jpg 61 125 | RealWorld/ToothBrush/00053.jpg 61 126 | RealWorld/Toys/00007.jpg 62 127 | RealWorld/Toys/00041.jpg 62 128 | RealWorld/Trash_Can/00029.jpg 63 129 | RealWorld/Trash_Can/00065.jpg 63 130 | RealWorld/Webcam/00012.jpg 64 131 | RealWorld/Webcam/00043.jpg 64 132 | -------------------------------------------------------------------------------- /data/splits/office_home/Real_labeled_p03.txt: -------------------------------------------------------------------------------- 1 | Real/Alarm_Clock/00001.jpg 0 2 | Real/Alarm_Clock/00015.jpg 0 3 | Real/Alarm_Clock/00057.jpg 0 4 | Real/Backpack/00005.jpg 1 5 | Real/Backpack/00012.jpg 1 6 | Real/Batteries/00011.jpg 2 7 | Real/Batteries/00013.jpg 2 8 | Real/Bed/00041.jpg 3 9 | Real/Bed/00054.jpg 3 10 | Real/Bike/00003.jpg 4 11 | Real/Bike/00004.jpg 4 12 | Real/Bike/00045.jpg 4 13 | Real/Bottle/00016.jpg 5 14 | Real/Bottle/00062.jpg 5 15 | Real/Bucket/00012.jpg 6 16 | Real/Bucket/00020.jpg 6 17 | Real/Calculator/00015.jpg 7 18 | Real/Calculator/00024.jpg 7 19 | Real/Calendar/00002.jpg 8 20 | Real/Calendar/00065.jpg 8 21 | Real/Candles/00009.jpg 9 22 | Real/Candles/00032.jpg 9 23 | Real/Candles/00063.jpg 9 24 | Real/Chair/00060.jpg 10 25 | Real/Chair/00084.jpg 10 26 | Real/Chair/00086.jpg 10 27 | Real/Clipboards/00009.jpg 11 28 | Real/Clipboards/00024.jpg 11 29 | Real/Computer/00006.jpg 12 30 | Real/Computer/00035.jpg 12 31 | Real/Couch/00003.jpg 13 32 | Real/Couch/00059.jpg 13 33 | Real/Curtains/00039.jpg 14 34 | Real/Curtains/00062.jpg 14 35 | Real/Desk_Lamp/00023.jpg 15 36 | Real/Desk_Lamp/00055.jpg 15 37 | Real/Drill/00016.jpg 16 38 | Real/Drill/00018.jpg 16 39 | Real/Eraser/00014.jpg 17 40 | Real/Eraser/00035.jpg 17 41 | Real/Exit_Sign/00059.jpg 18 42 | Real/Exit_Sign/00068.jpg 18 43 | Real/Fan/00002.jpg 19 44 | Real/Fan/00043.jpg 19 45 | Real/File_Cabinet/00011.jpg 20 46 | Real/File_Cabinet/00033.jpg 20 47 | Real/Flipflops/00019.jpg 21 48 | Real/Flipflops/00043.jpg 21 49 | Real/Flowers/00049.jpg 22 50 | Real/Flowers/00061.jpg 22 51 | Real/Folder/00004.jpg 23 52 | Real/Folder/00021.jpg 23 53 | Real/Fork/00010.jpg 24 54 | Real/Glasses/00032.jpg 25 55 | Real/Glasses/00038.jpg 25 56 | Real/Hammer/00005.jpg 26 57 | Real/Hammer/00011.jpg 26 58 | Real/Helmet/00014.jpg 27 59 | Real/Helmet/00036.jpg 27 60 | Real/Kettle/00004.jpg 28 61 | Real/Kettle/00019.jpg 28 62 | Real/Keyboard/00015.jpg 29 63 | Real/Keyboard/00045.jpg 29 64 | Real/Knives/00019.jpg 30 65 | Real/Knives/00068.jpg 30 66 | Real/Lamp_Shade/00040.jpg 31 67 | Real/Lamp_Shade/00059.jpg 31 68 | Real/Laptop/00006.jpg 32 69 | Real/Laptop/00054.jpg 32 70 | Real/Marker/00008.jpg 33 71 | Real/Monitor/00006.jpg 34 72 | Real/Monitor/00023.jpg 34 73 | Real/Mop/00008.jpg 35 74 | Real/Mouse/00010.jpg 36 75 | Real/Mouse/00020.jpg 36 76 | Real/Mug/00010.jpg 37 77 | Real/Mug/00055.jpg 37 78 | Real/Notebook/00006.jpg 38 79 | Real/Notebook/00060.jpg 38 80 | Real/Oven/00014.jpg 39 81 | Real/Oven/00042.jpg 39 82 | Real/Pan/00016.jpg 40 83 | Real/Paper_Clip/00044.jpg 41 84 | Real/Paper_Clip/00060.jpg 41 85 | Real/Pen/00006.jpg 42 86 | Real/Pen/00061.jpg 42 87 | Real/Pencil/00031.jpg 43 88 | Real/Pencil/00051.jpg 43 89 | Real/Postit_Notes/00010.jpg 44 90 | Real/Postit_Notes/00041.jpg 44 91 | Real/Printer/00047.jpg 45 92 | Real/Printer/00050.jpg 45 93 | Real/Push_Pin/00034.jpg 46 94 | Real/Push_Pin/00053.jpg 46 95 | Real/Radio/00038.jpg 47 96 | Real/Radio/00045.jpg 47 97 | Real/Refrigerator/00019.jpg 48 98 | Real/Refrigerator/00041.jpg 48 99 | Real/Ruler/00012.jpg 49 100 | Real/Scissors/00054.jpg 50 101 | Real/Scissors/00056.jpg 50 102 | Real/Screwdriver/00008.jpg 51 103 | Real/Screwdriver/00050.jpg 51 104 | Real/Shelf/00014.jpg 52 105 | Real/Shelf/00063.jpg 52 106 | Real/Sink/00013.jpg 53 107 | Real/Sink/00064.jpg 53 108 | Real/Sneakers/00025.jpg 54 109 | Real/Sneakers/00045.jpg 54 110 | Real/Sneakers/00083.jpg 54 111 | Real/Soda/00007.jpg 55 112 | Real/Soda/00045.jpg 55 113 | Real/Speaker/00021.jpg 56 114 | Real/Speaker/00075.jpg 56 115 | Real/Spoon/00010.jpg 57 116 | Real/Spoon/00018.jpg 57 117 | Real/TV/00021.jpg 58 118 | Real/TV/00047.jpg 58 119 | Real/Table/00018.jpg 59 120 | Real/Table/00054.jpg 59 121 | Real/Telephone/00010.jpg 60 122 | Real/Telephone/00011.jpg 60 123 | Real/ToothBrush/00011.jpg 61 124 | Real/ToothBrush/00030.jpg 61 125 | Real/ToothBrush/00053.jpg 61 126 | Real/Toys/00007.jpg 62 127 | Real/Toys/00041.jpg 62 128 | Real/Trash_Can/00029.jpg 63 129 | Real/Trash_Can/00065.jpg 63 130 | Real/Webcam/00012.jpg 64 131 | Real/Webcam/00043.jpg 64 132 | -------------------------------------------------------------------------------- /lib/iccv-poster.pdf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bostoncake/C-VisDiT/fb2dee61c874c3803f9bca5c7c4253a8b17e6230/lib/iccv-poster.pdf -------------------------------------------------------------------------------- /lib/overview.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bostoncake/C-VisDiT/fb2dee61c874c3803f9bca5c7c4253a8b17e6230/lib/overview.png -------------------------------------------------------------------------------- /lib/result1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bostoncake/C-VisDiT/fb2dee61c874c3803f9bca5c7c4253a8b17e6230/lib/result1.png -------------------------------------------------------------------------------- /lib/result2.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bostoncake/C-VisDiT/fb2dee61c874c3803f9bca5c7c4253a8b17e6230/lib/result2.png -------------------------------------------------------------------------------- /src/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Bostoncake/C-VisDiT/fb2dee61c874c3803f9bca5c7c4253a8b17e6230/src/__init__.py -------------------------------------------------------------------------------- /src/agents/__init__.py: -------------------------------------------------------------------------------- 1 | from .BaseAgent import BaseAgent 2 | from .CVisDiT import CVisDiT -------------------------------------------------------------------------------- /src/models/__init__.py: -------------------------------------------------------------------------------- 1 | from .clustering import compute_variance, torch_kmeans 2 | from .head import CosineClassifier 3 | from .memorybank import MemoryBank 4 | from .ssda import SSDALossModule, loss_info, update_data_memory 5 | -------------------------------------------------------------------------------- /src/models/clustering.py: -------------------------------------------------------------------------------- 1 | import logging 2 | import time 3 | from collections import Counter 4 | 5 | import faiss 6 | import numpy as np 7 | import torch 8 | import torch.nn as nn 9 | import torch.nn.functional as F 10 | import torchvision 11 | from faiss import Kmeans as faiss_Kmeans 12 | from tqdm import tqdm 13 | 14 | DEFAULT_KMEANS_SEED = 1234 15 | 16 | 17 | class Kmeans(object): 18 | def __init__( 19 | self, k_list, data, epoch=0, init_centroids=None, frozen_centroids=False 20 | ): 21 | """ 22 | Performs many k-means clustering. 23 | 24 | Args: 25 | data (np.array N * dim): data to cluster 26 | """ 27 | super().__init__() 28 | self.k_list = k_list 29 | self.data = data 30 | self.d = data.shape[-1] 31 | self.init_centroids = init_centroids 32 | self.frozen_centroids = frozen_centroids 33 | 34 | self.logger = logging.getLogger("Kmeans") 35 | self.debug = False 36 | self.epoch = epoch + 1 37 | 38 | def compute_clusters(self): 39 | """compute cluster 40 | 41 | Returns: 42 | torch.tensor, list: clus_labels, centroids 43 | """ 44 | data = self.data 45 | labels = [] 46 | centroids = [] 47 | 48 | tqdm_batch = tqdm(total=len(self.k_list), desc="[K-means]") 49 | for k_idx, each_k in enumerate(self.k_list): 50 | seed = k_idx * self.epoch + DEFAULT_KMEANS_SEED 51 | kmeans = faiss_Kmeans( 52 | self.d, 53 | each_k, 54 | niter=40, 55 | verbose=False, 56 | spherical=True, 57 | min_points_per_centroid=1, 58 | max_points_per_centroid=10000, 59 | gpu=True, 60 | seed=seed, 61 | frozen_centroids=self.frozen_centroids, 62 | ) 63 | 64 | kmeans.train(data, init_centroids=self.init_centroids) 65 | 66 | _, I = kmeans.index.search(data, 1) 67 | labels.append(I.squeeze(1)) 68 | C = kmeans.centroids 69 | centroids.append(C) 70 | 71 | tqdm_batch.update() 72 | tqdm_batch.close() 73 | 74 | labels = np.stack(labels, axis=0) 75 | 76 | return labels, centroids 77 | 78 | 79 | def torch_kmeans(k_list, data, init_centroids=None, seed=0, frozen=False): 80 | if init_centroids is not None: 81 | init_centroids = init_centroids.cpu().numpy() 82 | km = Kmeans( 83 | k_list, 84 | data.cpu().detach().numpy(), 85 | epoch=seed, 86 | frozen_centroids=frozen, 87 | init_centroids=init_centroids, 88 | ) 89 | clus_labels, centroids_npy = km.compute_clusters() 90 | clus_labels = torch.from_numpy(clus_labels).long().cuda() 91 | centroids = [] 92 | for c in centroids_npy: 93 | centroids.append(torch.from_numpy(c).cuda()) 94 | # compute phi 95 | clus_phi = [] 96 | for i in range(len(k_list)): 97 | clus_phi.append(compute_variance(data, clus_labels[i], centroids[i])) 98 | 99 | return clus_labels, centroids, clus_phi 100 | 101 | 102 | # variance 103 | 104 | 105 | @torch.no_grad() 106 | def compute_variance( 107 | data, cluster_labels, centroids, alpha=10, debug=False, num_class=None 108 | ): 109 | """compute variance for proto 110 | 111 | Args: 112 | data (torch.Tensor): data with shape [n, dim] 113 | cluster_labels (torch.Tensor): cluster labels of [n] 114 | centroids (torch.Tensor): cluster centroids [k, ndim] 115 | alpha (int, optional): Defaults to 10. 116 | debug (bool, optional): Defaults to False. 117 | 118 | Returns: 119 | [type]: [description] 120 | """ 121 | 122 | k = len(centroids) if num_class is None else num_class 123 | phis = torch.zeros(k) 124 | for c in range(k): 125 | cluster_points = data[cluster_labels == c] 126 | c_len = len(cluster_points) 127 | if c_len == 0: 128 | phis[c] = -1 129 | elif c_len == 1: 130 | phis[c] = 0.05 131 | else: 132 | phis[c] = torch.sum(torch.norm(cluster_points - centroids[c], dim=1)) / ( 133 | c_len * np.log(c_len + alpha) 134 | ) 135 | if phis[c] < 0.05: 136 | phis[c] = 0.05 137 | 138 | if debug: 139 | print("size-phi:", end=" ") 140 | for i in range(k): 141 | size = (cluster_labels == i).sum().item() 142 | print(f"{size}[phi={phis[i].item():.3f}]", end=", ") 143 | print("\n") 144 | 145 | return phis 146 | -------------------------------------------------------------------------------- /src/models/head.py: -------------------------------------------------------------------------------- 1 | import torch 2 | import torch.nn as nn 3 | import torch.nn.functional as F 4 | from utils.torchutils import grad_reverse 5 | from torch.autograd import Function 6 | from torchvision import models 7 | 8 | 9 | class CosineClassifier(nn.Module): 10 | def __init__(self, num_class=64, inc=4096, temp=0.05): 11 | super(CosineClassifier, self).__init__() 12 | self.fc = nn.Linear(inc, num_class, bias=False) 13 | self.num_class = num_class 14 | self.temp = temp 15 | 16 | def forward(self, x, reverse=False, eta=0.1): 17 | self.normalize_fc() 18 | 19 | if reverse: 20 | x = grad_reverse(x, eta) 21 | x = F.normalize(x) 22 | x_out = self.fc(x) 23 | x_out = x_out / self.temp 24 | 25 | return x_out 26 | 27 | def normalize_fc(self): 28 | self.fc.weight.data = F.normalize(self.fc.weight.data, p=2, eps=1e-12, dim=1) 29 | 30 | @torch.no_grad() 31 | def compute_discrepancy(self): 32 | self.normalize_fc() 33 | W = self.fc.weight.data 34 | D = torch.mm(W, W.transpose(0, 1)) 35 | D_mask = 1 - torch.eye(self.num_class).cuda() 36 | return torch.sum(D * D_mask).item() 37 | -------------------------------------------------------------------------------- /src/models/memorybank.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | import torch 3 | import torch.nn as nn 4 | import torch.nn.functional as F 5 | import torchvision 6 | 7 | 8 | class MemoryBank(object): 9 | """For efficiently computing the background vectors.""" 10 | 11 | def __init__(self, size, dim): 12 | """generate random memory bank 13 | 14 | Args: 15 | size (int): length of the memory bank 16 | dim (int): dimension of the memory bank features 17 | device_ids (list): gpu lists 18 | """ 19 | self._bank = self._create(size, dim) 20 | 21 | def _create(self, size, dim): 22 | """generate randomized features 23 | """ 24 | # initialize random weights 25 | mb_init = torch.rand(size, dim, device=torch.device("cuda")) 26 | std_dev = 1.0 / np.sqrt(dim / 3) 27 | mb_init = mb_init * (2 * std_dev) - std_dev 28 | # L2 normalize so that the norm is 1 29 | mb_init = F.normalize(mb_init) 30 | return mb_init.detach() # detach so its not trainable 31 | 32 | def as_tensor(self): 33 | return self._bank 34 | 35 | def at_idxs(self, idxs): 36 | return torch.index_select(self._bank, 0, idxs) 37 | 38 | def update(self, indices, data_memory): 39 | data_dim = data_memory.size(1) 40 | data_memory = data_memory.detach() 41 | indices = indices.unsqueeze(1).repeat(1, data_dim) 42 | 43 | self._bank = self._bank.scatter_(0, indices, data_memory) 44 | -------------------------------------------------------------------------------- /src/utils/__init__.py: -------------------------------------------------------------------------------- 1 | from . import datautils, torchutils, utils 2 | from .setup import check_pretrain_dir, print_info, process_config 3 | from .utils import * 4 | -------------------------------------------------------------------------------- /src/utils/setup.py: -------------------------------------------------------------------------------- 1 | import datetime 2 | import logging 3 | import os 4 | import shutil 5 | import socket 6 | from logging import Formatter 7 | from logging.handlers import RotatingFileHandler 8 | from pprint import pprint 9 | from tempfile import mkstemp 10 | 11 | from dotmap import DotMap 12 | 13 | from .utils import load_json, makedirs, save_json 14 | 15 | 16 | def check_pretrain_dir(config_json): 17 | pre_checkpoint_dir = None 18 | if ( 19 | "pretrained_exp_dir" in config_json 20 | and 21 | config_json["pretrained_exp_dir"] is not None 22 | ): 23 | print("NOTE: found pretrained model...continue training") 24 | pre_checkpoint_dir = os.path.join( 25 | config_json["pretrained_exp_dir"], "checkpoints" 26 | ) 27 | return pre_checkpoint_dir 28 | 29 | 30 | def process_config_path(config_path, override_dotmap=None): 31 | config_json = load_json(config_path) 32 | return process_config(config_json, override_dotmap=override_dotmap) 33 | 34 | 35 | def process_config(config_json, override_dotmap=None): 36 | """ 37 | Processes config file: 38 | 1) Converts it to a DotMap 39 | 2) Creates experiments path and required subdirs 40 | 3) Set up logging 41 | """ 42 | config = DotMap(config_json) 43 | if override_dotmap is not None: 44 | config.update(override_dotmap) 45 | 46 | print("Configuration Loaded:") 47 | pprint(config) 48 | 49 | print() 50 | print(" *********************************************** ") 51 | print(" Running C-VisDiT on {} benchmark".format(config.exp_name)) 52 | print(" *********************************************** ") 53 | print() 54 | 55 | # if config.pretrained_exp_dir is not None: 56 | # # don't make new dir more continuing training 57 | # exp_dir = config.pretrained_exp_dir 58 | # print("[INFO]: Continuing from previously finished training at %s." % exp_dir) 59 | # else: 60 | exp_base = config.exp_base 61 | 62 | if config.debug: 63 | exp_dir = os.path.join(exp_base, "experiments", config.exp_name, "debug") 64 | else: 65 | if config.pretrained_exp_dir is not None and isinstance( 66 | config.pretrained_exp_dir, str 67 | ): 68 | # don't make new dir more continuing training 69 | exp_dir = config.pretrained_exp_dir 70 | print("[INFO]: Backup previously trained model and config json") 71 | os.system("cp %s/config.json %s/prev_config.json" % (exp_dir, exp_dir)) 72 | os.system( 73 | "cp %s/checkpoints/checkpoint.pth.tar %s/checkpoints/prev_checkpoint.pth.tar" 74 | % (exp_dir, exp_dir) 75 | ) 76 | os.system( 77 | "cp %s/checkpoints/model_best.pth.tar %s/checkpoints/prev_model_best.pth.tar" 78 | % (exp_dir, exp_dir) 79 | ) 80 | elif config.continue_exp_dir is not None and isinstance( 81 | config.continue_exp_dir, str 82 | ): 83 | exp_dir = config.continue_exp_dir 84 | print("[INFO]: Backup previously trained model and config json") 85 | os.system("cp %s/config.json %s/prev_config.json" % (exp_dir, exp_dir)) 86 | os.system( 87 | "cp %s/checkpoints/checkpoint.pth.tar %s/checkpoints/prev_checkpoint.pth.tar" 88 | % (exp_dir, exp_dir) 89 | ) 90 | os.system( 91 | "cp %s/checkpoints/model_best.pth.tar %s/checkpoints/prev_model_best.pth.tar" 92 | % (exp_dir, exp_dir) 93 | ) 94 | else: 95 | if config.exp_id is None: 96 | config.exp_id = datetime.datetime.now().strftime("%Y-%m-%d") 97 | exp_dir = os.path.join( 98 | exp_base, "experiments", config.exp_name, config.exp_id 99 | ) 100 | if os.path.exists(exp_dir): 101 | config.exp_id += "-" + datetime.datetime.now().strftime("%y%m%d%H%M%S") 102 | exp_dir = os.path.join( 103 | exp_base, "experiments", config.exp_name, config.exp_id 104 | ) 105 | 106 | # create some important directories to be used for the experiment. 107 | config.summary_dir = os.path.join(exp_dir, "summaries/") 108 | config.checkpoint_dir = os.path.join(exp_dir, "checkpoints/") 109 | config.out_dir = os.path.join(exp_dir, "out/") 110 | config.log_dir = os.path.join(exp_dir, "logs/") 111 | 112 | makedirs( 113 | [config.summary_dir, config.checkpoint_dir, config.out_dir, config.log_dir] 114 | ) 115 | 116 | # save config to experiment dir 117 | config_out = os.path.join(exp_dir, "config.json") 118 | save_json(config.toDict(), config_out) 119 | 120 | # setup logging in the project 121 | setup_logging(config.log_dir) 122 | 123 | logging.getLogger().info("Experiment directory is located at %s" % exp_dir) 124 | 125 | logging.getLogger().info("Configurations and directories successfully set up.") 126 | return config 127 | 128 | 129 | def setup_logging(log_dir): 130 | log_file_format = ( 131 | "[%(levelname)s] %(asctime)s: %(message)s in %(pathname)s:%(lineno)d" 132 | ) 133 | log_console_format = "[%(levelname)s]: %(message)s" 134 | 135 | # Main logger 136 | main_logger = logging.getLogger() 137 | main_logger.setLevel(logging.INFO) 138 | 139 | console_handler = logging.StreamHandler() 140 | console_handler.setLevel(logging.INFO) 141 | console_handler.setFormatter(Formatter(log_console_format)) 142 | 143 | exp_file_handler = RotatingFileHandler( 144 | "{}exp_debug.log".format(log_dir), maxBytes=10 ** 6, backupCount=5 145 | ) 146 | exp_file_handler.setLevel(logging.DEBUG) 147 | exp_file_handler.setFormatter(Formatter(log_file_format)) 148 | 149 | exp_errors_file_handler = RotatingFileHandler( 150 | "{}exp_error.log".format(log_dir), maxBytes=10 ** 6, backupCount=5 151 | ) 152 | exp_errors_file_handler.setLevel(logging.WARNING) 153 | exp_errors_file_handler.setFormatter(Formatter(log_file_format)) 154 | 155 | main_logger.addHandler(console_handler) 156 | main_logger.addHandler(exp_file_handler) 157 | main_logger.addHandler(exp_errors_file_handler) 158 | 159 | 160 | def print_info(output=print): 161 | output(f"Start at time: {datetime.datetime.now().strftime('%Y.%m.%d-%H:%M:%S')}") 162 | output(f"Server: {socket.gethostname()}") 163 | 164 | 165 | def prepare_dirs(config): 166 | exp_dir = os.path.join( 167 | config.exp_base, "experiments", config.exp_name, config.exp_id 168 | ) 169 | summary_dir = os.path.join(exp_dir, "summaries/") 170 | checkpoint_dir = os.path.join(exp_dir, "checkpoints/") 171 | out_dir = os.path.join(exp_dir, "out/") 172 | log_dir = os.path.join(exp_dir, "logs/") 173 | config.log_file = os.path.join(log_dir, "output.log") 174 | if config.pretrained_exp_dir == None or config.copy_exp_dir is False: 175 | makedirs([summary_dir, checkpoint_dir, out_dir, log_dir]) 176 | print(f"Create {exp_dir}") 177 | else: 178 | shutil.copytree(config.pretrained_exp_dir, exp_dir) 179 | if os.path.exists(config.log_file): 180 | shutil.copy(config.log_file, os.path.join(log_dir, "output_prev.log")) 181 | print(f"Copy {config.pretrained_exp_dir} to {exp_dir}") 182 | config.pretrained_exp_dir = exp_dir 183 | 184 | 185 | def get_cmd( 186 | config, script_path="/rscratch/xyyue/anaconda3/envs/ssda2/bin/python ./run.py" 187 | ): 188 | config_out = mkstemp()[1] 189 | save_json(config.toDict(), config_out) 190 | return f"{script_path} --config {config_out}" 191 | -------------------------------------------------------------------------------- /src/utils/utils.py: -------------------------------------------------------------------------------- 1 | import gc 2 | import json 3 | import os 4 | import random 5 | import shutil 6 | import string 7 | from collections import Counter, OrderedDict 8 | 9 | import numpy as np 10 | import torch 11 | from dotmap import DotMap 12 | 13 | # Specific 14 | 15 | DOMAIN_V = {"source": "target", "target": "source"} 16 | 17 | 18 | def reverse_domain(domain_name): 19 | return DOMAIN_V[domain_name] 20 | 21 | 22 | def per(acc): 23 | return f"{acc * 100:.2f}%" 24 | 25 | 26 | # Debug 27 | 28 | 29 | def MB(x): 30 | return f"{x/1024/1024:.3} MB" 31 | 32 | 33 | def GB(x): 34 | return f"{x/1024/1024/1024:.3} GB" 35 | 36 | 37 | def print_occupied_mem(idx=0): 38 | print(GB(torch.cuda.memory_allocated(idx))) 39 | 40 | 41 | def size_of_tensor(x): 42 | return MB(x.element_size() * x.nelement()) 43 | 44 | 45 | def randtext(length=10): 46 | return "".join([random.choice(string.ascii_letters) for i in range(length)]) 47 | 48 | 49 | # OS 50 | 51 | 52 | def to_list(something): 53 | if something is not None and not isinstance(something, list): 54 | return [something] 55 | return something 56 | 57 | 58 | def makedirs(dir_list): 59 | if not isinstance(dir_list, list): 60 | dir_list = [dir_list] 61 | for dir in dir_list: 62 | if not os.path.exists(dir): 63 | os.makedirs(dir) 64 | 65 | 66 | # Counter 67 | 68 | 69 | class AverageMeter(object): 70 | """Computes and stores the average and current value""" 71 | 72 | def __init__(self): 73 | self.reset() 74 | 75 | def reset(self): 76 | self.val = 0 77 | self.avg = 0 78 | self.sum = 0 79 | self.count = 0 80 | 81 | def update(self, val, n=1): 82 | self.val = val 83 | self.sum += val * n 84 | self.count += n 85 | self.avg = self.sum / self.count 86 | 87 | 88 | class ProgressMeter(object): 89 | def __init__(self, num_batches, meters, prefix=""): 90 | self.batch_fmtstr = self._get_batch_fmtstr(num_batches) 91 | self.meters = meters 92 | self.prefix = prefix 93 | 94 | def display(self, batch): 95 | entries = [self.prefix + self.batch_fmtstr.format(batch)] 96 | entries += [str(meter) for meter in self.meters] 97 | print("\t".join(entries)) 98 | 99 | def _get_batch_fmtstr(self, num_batches): 100 | num_digits = len(str(num_batches // 1)) 101 | fmt = "{:" + str(num_digits) + "d}" 102 | return "[" + fmt + "/" + fmt.format(num_batches) + "]" 103 | 104 | 105 | class OrderedCounter(Counter, OrderedDict): 106 | """Counter that remembers the order elements are first encountered""" 107 | 108 | def __repr__(self): 109 | return "%s(%r)" % (self.__class__.__name__, OrderedDict(self)) 110 | 111 | def __reduce__(self): 112 | return self.__class__, (OrderedDict(self),) 113 | 114 | 115 | # Json 116 | 117 | 118 | def load_json(f_path): 119 | with open(f_path, "r") as f: 120 | return json.load(f) 121 | 122 | 123 | def save_json(obj, f_path): 124 | with open(f_path, "w") as f: 125 | json.dump(obj, f, ensure_ascii=False, indent=4) 126 | 127 | 128 | def is_div(freq, epoch, best=False): 129 | if freq is not None and best: 130 | return True 131 | return freq and epoch % freq == 0 132 | 133 | 134 | def info_gpu_usage(): 135 | for obj in gc.get_objects(): 136 | try: 137 | if torch.is_tensor(obj) or ( 138 | hasattr(obj, "data") and torch.is_tensor(obj.data) 139 | ): 140 | print(type(obj), obj.size()) 141 | except: 142 | pass 143 | 144 | 145 | # dotmap 146 | 147 | 148 | def exist_key(k): 149 | is_empty_dotmap = isinstance(k, DotMap) and len(k) == 0 150 | return isinstance(k, bool) or (not is_empty_dotmap and k is not None) 151 | 152 | 153 | def set_default(cur_config, name, value=None, callback=None): 154 | if not exist_key(cur_config[name]): 155 | if value is not None: 156 | cur_config[name] = value 157 | elif callback is not None: 158 | assert exist_key(cur_config[callback]) 159 | cur_config[name] = cur_config[callback] 160 | elif value is None and callback is None: 161 | cur_config[name] = value 162 | else: 163 | raise NotImplementedError 164 | return cur_config[name] 165 | -------------------------------------------------------------------------------- /train.sh: -------------------------------------------------------------------------------- 1 | currenttime=`date "+%Y%m%d_%H%M%S"` 2 | config=config/office/A-D-1.json # choose the corresponding configurations in `./config` 3 | exp_id=${currenttime}"_Office-31_Amazon-DSLR-1-shot" # rename the experiment folder to distinguish between experiments 4 | CUDA_VISIBLE_DEVICES=0 python src/run.py \ 5 | --config ${config} \ 6 | --exp_id ${exp_id} 7 | --------------------------------------------------------------------------------