└── README.md /README.md: -------------------------------------------------------------------------------- 1 | ## Awesome Source-free Test-time Adaptation   [![Awesome](https://cdn.rawgit.com/sindresorhus/awesome/d7305f38d29fed78fa85652e3a63e154dd8e8829/media/badge.svg)](https://github.com/sindresorhus/awesome) [![Maintenance](https://img.shields.io/badge/Maintained%3F-yes-green.svg)](https://GitHub.com/Naereen/StrapDown.js/graphs/commit-activity) 2 | 3 | This is a curated list of research papers in `Test-time Adaptation` (**TTA**), which also goes by other names, such as `Test-time Training` (**TTT**), `Source-free Domain Adaptation` (**SFDA**) and `Unsupervised Model Adaptation` (**UMA**). 4 | 5 | The repository is actively maintained. Pull requests or direct messages are welcome. 6 | 7 | ### Table of Contents 8 | 9 | * [Methods](#methods) 10 | * [Self-supervision](#self-supervision) 11 | * [Information Entropy](#information-entropy) 12 | * [Batch Normalization](#batch-normalization) 13 | * [Pseudo Labeling](#pseudo-labeling) 14 | * [Class Prototype](#class-prototype) 15 | * [Feature Alignment](#feature-alignment) 16 | * [Generative Modeling](#generative-modeling) 17 | * [Nearest Neighbors](#nearest-neighbors) 18 | * [Augmentation Invariance](#augmentation-invariance) 19 | * [Meta-learning](#meta-learning) 20 | * [Time-varying](#time-varying) 21 | * [Others](#others) 22 | * [Benchmark](#benchmark) 23 | * [Applications](#applications) 24 | 25 | ### Methods 26 | 27 | #### Self-supervision 28 | 29 | * [Test-Time Training with Self-Supervision for Generalization under Distribution Shifts](http://proceedings.mlr.press/v119/sun20b.html) ICML'20 [[Project]](https://yueatsprograms.github.io/ttt/home.html) 30 | * [TTT++: When Does Self-Supervised Test-Time Training Fail or Thrive?](https://openreview.net/pdf?id=86NHK__yFDl) NeurIPS'21 [[Code]](https://github.com/vita-epfl/ttt-plus-plus) 31 | * [Model Adaptation: Historical Contrastive Learning for Unsupervised Domain Adaptation without Source Data](https://openreview.net/pdf?id=0zXJRJecC_) NeurIPS'21 32 | * [Contrastive Test-Time Adaptation](https://openaccess.thecvf.com/content/CVPR2022/papers/Chen_Contrastive_Test-Time_Adaptation_CVPR_2022_paper.pdf) CVPR'22 [[Code]](https://github.com/DianCh/AdaContrast) 33 | * [Divide and Contrast: Source-free Domain Adaptation via Adaptive Contrastive Learning](https://openreview.net/forum?id=NjImFaBEHl) NeurIPS'22 34 | * [Test-Time Training with Masked Autoencoders](https://openreview.net/forum?id=SHMi1b7sjXk) NeurIPS'22 [[Project]](https://yossigandelsman.github.io/ttt_mae/index.html) 35 | * [Improved Test-Time Adaptation for Domain Generalization](https://arxiv.org/abs/2304.04494) CVPR'23 [[Code]](https://github.com/liangchen527/ITTA) 36 | * [PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalization](https://arxiv.org/abs/2307.15199) ICCV'23 37 | * [MATE: Masked Autoencoders are Online 3D Test-Time Learners](https://arxiv.org/abs/2211.11432) ICCV'23 38 | * [On the Robustness of Open-World Test-Time Training: Self-Training with Dynamic Prototype Expansion](https://arxiv.org/abs/2308.09942) ICCV'23 [[Code]](https://github.com/Yushu-Li/OWTTT) 39 | 40 | #### Information Entropy 41 | 42 | * [Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation](http://proceedings.mlr.press/v119/liang20a.html) ICML'20 [[Code]](https://github.com/tim-learn/SHOT) 43 | * [Tent: Fully Test-Time Adaptation by Entropy Minimization](https://openreview.net/forum?id=uXl3bZLkr3c) ICLR'21 [[Code]](https://github.com/DequanWang/tent) 44 | * [Uncertainty Reduction for Model Adaptation in Semantic Segmentation](https://openaccess.thecvf.com/content/CVPR2021/html/S_Uncertainty_Reduction_for_Model_Adaptation_in_Semantic_Segmentation_CVPR_2021_paper.html) CVPR'21 [[Code]](https://github.com/idiap/model-uncertainty-for-adaptation) 45 | * [Bayesian Adaptation for Covariate Shift](https://openreview.net/forum?id=15HPeY8MGQ) NeurIPS'21 46 | * [Efficient Test-Time Model Adaptation without Forgetting](https://arxiv.org/abs/2204.02610) ICML'22 [[Code]](https://github.com/mr-eggplant/EATA) 47 | * [Confidence Score for Source-Free Unsupervised Domain Adaptation](https://arxiv.org/abs/2206.06640) ICML'22 [[Code]](https://github.com/Jhyun17/CoWA-JMDS) 48 | * [Towards Stable Test-time Adaptation in Dynamic Wild World](https://openreview.net/forum?id=g2YraF75Tj) ICLR'23 [[Code]](https://github.com/mr-eggplant/SAR) 49 | 50 | #### Batch Normalization 51 | 52 | * [Improving robustness against common corruptions by covariate shift adaptation](https://proceedings.neurips.cc/paper/2020/file/85690f81aadc1749175c187784afc9ee-Paper.pdf) NeurIPS'20 [[Code]](https://github.com/bethgelab/robustness) 53 | * [Tent: Fully Test-Time Adaptation by Entropy Minimization](https://openreview.net/forum?id=uXl3bZLkr3c) ICLR'21 [[Code]](https://github.com/DequanWang/tent) 54 | * [Limitations of Post-Hoc Feature Alignment for Robustness 55 | ](https://openaccess.thecvf.com/content/CVPR2021/html/Burns_Limitations_of_Post-Hoc_Feature_Alignment_for_Robustness_CVPR_2021_paper.html) CVPR'21 [[Code]](https://github.com/collin-burns/feature-alignment) 56 | * [TTN: A Domain-Shift Aware Batch Normalization in Test-Time Adaptation](https://openreview.net/forum?id=EQfeudmWLQ) ICLR'23 57 | * [Delta: Degradation-Free Fully Test-Time Adaptation](https://openreview.net/forum?id=eGm22rqG93) ICLR'23 58 | * [Towards Stable Test-time Adaptation in Dynamic Wild World](https://openreview.net/forum?id=g2YraF75Tj) ICLR'23 [[Code]](https://github.com/mr-eggplant/SAR) 59 | * [Test-time Adaptation in the Dynamic World with Compound Domain Knowledge Management](https://arxiv.org/abs/2212.08356) RA-L'23 60 | 61 | #### Pseudo Labeling 62 | 63 | * [Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation](http://proceedings.mlr.press/v119/liang20a.html) ICML'20 [[Code]](https://github.com/tim-learn/SHOT) 64 | * [Generative Pseudo-label Refinement for Unsupervised Domain Adaptation](https://www.computer.org/csdl/proceedings-article/wacv/2020/09093579/1jPbrqho82k) WACV'20 65 | * [A Free Lunch for Unsupervised Domain Adaptive Object Detection without Source Data](https://ojs.aaai.org/index.php/AAAI/article/view/17029) AAAI'21 66 | * [Uncertainty Reduction for Model Adaptation in Semantic Segmentation](https://openaccess.thecvf.com/content/CVPR2021/html/S_Uncertainty_Reduction_for_Model_Adaptation_in_Semantic_Segmentation_CVPR_2021_paper.html) CVPR'21 [[Code]](https://github.com/idiap/model-uncertainty-for-adaptation) 67 | * [Adapting ImageNet-scale models to complex distribution shifts with self-learning](https://arxiv.org/abs/2104.12928) TMLR'22 [[Code]](https://github.com/bethgelab/robustness) 68 | * [Continual Test-Time Domain Adaptation](https://arxiv.org/abs/2203.13591) CVPR'22 [[Code]](https://github.com/qinenergy/cotta) 69 | * [Contrastive Test-Time Adaptation](https://openaccess.thecvf.com/content/CVPR2022/papers/Chen_Contrastive_Test-Time_Adaptation_CVPR_2022_paper.pdf) CVPR'22 [[Code]](https://github.com/DianCh/AdaContrast) 70 | * [Test-Time Adaptation via Conjugate Pseudo-labels](https://openreview.net/forum?id=2yvUYc-YNUH) NeurIPS'22 [[Code]](https://github.com/locuslab/tta_conjugate) 71 | * [Towards Understanding GD with Hard and Conjugate Pseudo-labels for Test-Time Adaptation](https://openreview.net/forum?id=FJXf1FXN8C) ICLR'23 72 | * [TeSLA: Test-Time Self-Learning With Automatic Adversarial Augmentation](https://arxiv.org/abs/2303.09870) CVPR'23 [[Code]](https://github.com/devavratTomar/TeSLA) 73 | 74 | #### Class Prototype 75 | 76 | * [Model Adaptation: Unsupervised Domain Adaptation Without Source Data](https://openaccess.thecvf.com/content_CVPR_2020/html/Li_Model_Adaptation_Unsupervised_Domain_Adaptation_Without_Source_Data_CVPR_2020_paper.html) CVPR'20 77 | * [Test-Time Classifier Adjustment Module for Model-Agnostic Domain Generalization](https://proceedings.neurips.cc/paper/2021/hash/1415fe9fea0fa1e45dddcff5682239a0-Abstract.html) NeurIPS'21 [[Code]](https://github.com/matsuolab/T3A) 78 | * [Revisiting Realistic Test-Time Training: Sequential Inference and Adaptation by Anchored Clustering](https://openreview.net/forum?id=W-_4hgRkwb) NeurIPS'22 [[Code]](https://github.com/Gorilla-Lab-SCUT/TTAC) 79 | * [Attracting and Dispersing: A Simple Approach for Source-free Domain Adaptation](https://openreview.net/forum?id=ZlCpRiZN7n) NeurIPS'22 [[Code]](https://github.com/Albert0147/AaD_SFDA) 80 | * [On the Robustness of Open-World Test-Time Training: Self-Training with Dynamic Prototype Expansion](https://arxiv.org/abs/2308.09942) ICCV'23 [[Code]](https://github.com/Yushu-Li/OWTTT) 81 | 82 | #### Feature Alignment 83 | 84 | * [SoFA: Source-data-free Feature Alignment for Unsupervised Domain Adaptation](https://openaccess.thecvf.com/content/WACV2021/papers/Yeh_SoFA_Source-Data-Free_Feature_Alignment_for_Unsupervised_Domain_Adaptation_WACV_2021_paper.pdf) WACV'21 85 | * [Adaptive Adversarial Network for Source-Free Domain Adaptation](https://openaccess.thecvf.com/content/ICCV2021/html/Xia_Adaptive_Adversarial_Network_for_Source-Free_Domain_Adaptation_ICCV_2021_paper.html) ICCV'21 86 | * [TTT++: When Does Self-Supervised Test-Time Training Fail or Thrive?](https://proceedings.neurips.cc/paper/2021/hash/b618c3210e934362ac261db280128c22-Abstract.html) NeurIPS'21 [[Code]](https://github.com/vita-epfl/ttt-plus-plus) 87 | * [Source-Free Adaptation to Measurement Shift via Bottom-Up Feature Restoration](https://openreview.net/forum?id=1JDiK_TbV4S) ICLR'22 [[Code]](https://github.com/cianeastwood/bufr) 88 | * [Invariance Through Latent Alignment](https://openreview.net/forum?id=vXGcHthY6v) RSS'22 [[Project]](https://invariance-through-latent-alignment.github.io/) 89 | * [Source-Free Domain Adaptation via Distribution Estimation](https://openaccess.thecvf.com/content/CVPR2022/papers/Ding_Source-Free_Domain_Adaptation_via_Distribution_Estimation_CVPR_2022_paper.pdf) CVPR'22 90 | * [ActMAD: Activation Matching to Align Distributions for Test-Time-Training](https://arxiv.org/abs/2211.12870) CVPR'23 [[Code]](https://github.com/jmiemirza/actmad) 91 | * [Feature Alignment and Uniformity for Test Time Adaptation 92 | ](https://openaccess.thecvf.com/content/CVPR2023/html/Wang_Feature_Alignment_and_Uniformity_for_Test_Time_Adaptation_CVPR_2023_paper.html) CVPR'23 [[Code]](https://github.com/SakurajimaMaiii/TSD) 93 | * [Robustness to corruption in pre-trained Bayesian neural networks](https://openreview.net/forum?id=kUI41mY8bHl) ICLR'23 94 | 95 | #### Generative Modeling 96 | 97 | * [Model Adaptation: Unsupervised Domain Adaptation without Source Data](https://openaccess.thecvf.com/content_CVPR_2020/html/Li_Model_Adaptation_Unsupervised_Domain_Adaptation_Without_Source_Data_CVPR_2020_paper.html) CVPR'20 98 | * [Domain Impression: A Source Data Free Domain Adaptation Method](https://openaccess.thecvf.com/content/WACV2021/papers/Kurmi_Domain_Impression_A_Source_Data_Free_Domain_Adaptation_Method_WACV_2021_paper.pdf) WACV'21 99 | * [Back to the Source: Diffusion-Driven Test-Time Adaptation](https://arxiv.org/abs/2207.03442) CVPR'23 [[Code]](https://github.com/shiyegao/DDA) 100 | 101 | #### Nearest Neighbors 102 | 103 | * [Generalized Source-free Domain Adaptation](https://openaccess.thecvf.com/content/ICCV2021/papers/Yang_Generalized_Source-Free_Domain_Adaptation_ICCV_2021_paper.pdf) ICCV'21 [[Code]](https://github.com/Albert0147/G-SFDA) 104 | * [Exploiting the Intrinsic Neighborhood Structure for Source-free Domain Adaptation](https://proceedings.neurips.cc/paper/2021/hash/f5deaeeae1538fb6c45901d524ee2f98-Abstract.html) NeurIPS'21 [[Code]](https://github.com/Albert0147/NRC_SFDA) 105 | * [Test-Time Adaptation via Self-Training with Nearest Neighbor Information](https://openreview.net/forum?id=EzLtB4M1SbM) ICLR'23 [[Code]](https://github.com/mingukjang/TAST) 106 | 107 | #### Augmentation Invariance 108 | 109 | * [MEMO: Test Time Robustness via Adaptation and Augmentation](https://openreview.net/forum?id=vn74m_tWu8O) NeurIPS'22 [[Code]](https://github.com/zhangmarvin/memo) 110 | * [Test time Adaptation through Perturbation Robustness](https://openreview.net/forum?id=GbBeI5z86uD) NeurIPS-WS'21 111 | * [Balancing Discriminability and Transferability for Source-Free Domain Adaptation](https://arxiv.org/abs/2206.08009?context=cs.LG) ICML'22 [[Project]](https://sites.google.com/view/mixup-sfda) 112 | 113 | #### Meta-learning 114 | 115 | * [Test-Time Fast Adaptation for Dynamic Scene Deblurring via Meta-Auxiliary Learning](https://openaccess.thecvf.com/content/CVPR2021/html/Chi_Test-Time_Fast_Adaptation_for_Dynamic_Scene_Deblurring_via_Meta-Auxiliary_Learning_CVPR_2021_paper.html) CVPR'21 116 | * [Adaptive Risk Minimization: Learning to Adapt to Domain Shift](https://proceedings.neurips.cc/paper/2021/hash/c705112d1ec18b97acac7e2d63973424-Abstract.html) NeurIPS'21 [[Code]](https://sites.google.com/view/adaptive-risk-minimization) 117 | * [Learning to Generalize across Domains on Single Test Samples](https://openreview.net/forum?id=CIaQKbTBwtU) ICLR'22 [[Code]](https://github.com/zzzx1224/SingleSampleGeneralization-ICLR2022) 118 | * [Meta-DMoE: Adapting to Domain Shift by Meta-Distillation from Mixture-of-Experts](https://openreview.net/forum?id=_ekGcr07Dsp) NeurIPS'22 [[Code]](https://github.com/n3il666/Meta-DMoE) 119 | 120 | #### Time-varying 121 | * [Continual Test-Time Domain Adaptation](https://arxiv.org/abs/2203.13591) CVPR'22 [[Code]](https://github.com/qinenergy/cotta) 122 | * [NOTE: Robust Continual Test-time Adaptation Against Temporal Correlation](https://openreview.net/forum?id=E9HNxrCFZPV) NeurIPS'22 [[Code]](https://github.com/TaesikGong/NOTE) 123 | * [Extrapolative Continuous-time Bayesian Neural Network for Fast Training-free Test-time Adaptation](https://openreview.net/forum?id=wiHzQWwg3l) NeurIPS'22 [[Code]](https://github.com/guxm2021/ECBNN) 124 | * [Decorate the Newcomers: Visual Domain Prompt for Continual Test Time Adaptation](https://arxiv.org/pdf/2212.04145.pdf) AAAI'23 125 | * [Robust Test-Time Adaptation in Dynamic Scenarios](https://arxiv.org/abs/2303.13899) CVPR'23 [[Code]](https://github.com/BIT-DA/RoTTA) 126 | * [A Probabilistic Framework for Lifelong Test-Time Adaptation](https://arxiv.org/abs/2212.09713) CVPR'23 [[Code]](https://github.com/dhanajitb/petal) 127 | * [Robust Mean Teacher for Continual and Gradual Test-Time Adaptation](https://arxiv.org/abs/2211.13081) CVPR'23 [[Code]](https://github.com/mariodoebler/test-time-adaptation) 128 | * [EcoTTA: Memory-Efficient Continual Test-time Adaptation via Self-distilled Regularization](https://arxiv.org/abs/2303.01904) CVPR'23 [[Project]](https://sites.google.com/view/junha/ecotta) 129 | 130 | #### Others 131 | 132 | * [Domain Adaptation in the Absence of Source Domain Data](https://www.kdd.org/kdd2016/papers/files/adp0290-chidlovskiiA.pdf) KDD'16 133 | * [Semantic Photo Manipulation with a Generative Image Prior](https://arxiv.org/pdf/2005.07727.pdf) SIGGRAPH'19 134 | * [Collaborative Sampling in Generative Adversarial Networks](https://arxiv.org/pdf/1902.00813.pdf) AAAI'20 [[Code]](https://github.com/vita-epfl/collaborative-gan-sampling) 135 | * [Universal Source-Free Domain Adaptation](https://openaccess.thecvf.com/content_CVPR_2020/papers/Kundu_Universal_Source-Free_Domain_Adaptation_CVPR_2020_paper.pdf) CVPR'20 [[Project]](https://sites.google.com/view/usfda-cvpr2020) 136 | * [Adaptive Methods for Real-World Domain Generalization](https://openaccess.thecvf.com/content/CVPR2021/papers/Dubey_Adaptive_Methods_for_Real-World_Domain_Generalization_CVPR_2021_paper.pdf) CVPR'21 [[Code]](https://github.com/abhimanyudubey/PrototypicalDomainGeneralization) 137 | * [Parameter-free Online Test-time Adaptation](https://arxiv.org/abs/2201.05718) CVPR'22 [[Code]](https://github.com/fiveai/LAME) 138 | * [Evaluating the Adversarial Robustness of Adaptive Test-time Defenses](https://proceedings.mlr.press/v162/croce22a/croce22a.pdf) ICML'22 139 | * [MECTA: Memory-Economic Continual Test-Time Model Adaptation](https://openreview.net/forum?id=N92hjSf5NNh) ICLR'23 140 | * [Neuro-Modulated Hebbian Learning for Fully Test-Time Adaptation](https://arxiv.org/abs/2303.00914) CVPR'23 141 | 142 | ### Benchmark 143 | * [On Pitfalls of Test-Time Adaptation](https://arxiv.org/pdf/2306.03536.pdf) ICML'2023 [[Code]](https://github.com/LINs-lab/ttab) 144 | 145 | ### Applications 146 | * [Consistent Video Depth Estimation](https://arxiv.org/pdf/2004.15021.pdf) SIGGRAPH'2020 [[Project]](https://roxanneluo.github.io/Consistent-Video-Depth-Estimation/) 147 | * [Self-Supervised Policy Adaptation during Deployment](https://openreview.net/forum?id=o_V-MjyyGV_) ICLR'21 [[Project]](https://nicklashansen.github.io/PAD/) 148 | * [Source-Free Domain Adaptation for Image Segmentation](https://arxiv.org/pdf/2108.03152.pdf) MICCAI'20 [[Code]](https://github.com/mathilde-b/SFDA) 149 | * [Fully Test-Time Adaptation for Image Segmentation](https://link.springer.com/chapter/10.1007/978-3-030-87199-4_24) MICCAI'21 150 | * [Adapting Off-the-Shelf Source Segmenter for Target Medical Image Segmentation](https://arxiv.org/pdf/2106.12497.pdf) MICCAI'21 151 | * [Source-Free Domain Adaptation for Semantic Segmentation](https://openaccess.thecvf.com/content/CVPR2021/html/Liu_Source-Free_Domain_Adaptation_for_Semantic_Segmentation_CVPR_2021_paper.html) CVPR'21 152 | * [Generalize Then Adapt: Source-Free Domain Adaptive Semantic Segmentation 153 | ](https://openaccess.thecvf.com/content/ICCV2021/html/Kundu_Generalize_Then_Adapt_Source-Free_Domain_Adaptive_Semantic_Segmentation_ICCV_2021_paper.html) ICCV'21 [[Project]](https://sites.google.com/view/sfdaseg) 154 | * [SS-SFDA: Self-Supervised Source-Free Domain Adaptation for Road Segmentation in Hazardous Environments](https://openaccess.thecvf.com/content/WACV2021/papers/Yeh_SoFA_Source-Data-Free_Feature_Alignment_for_Unsupervised_Domain_Adaptation_WACV_2021_paper.pdf) ICCV'21 [[Project]](https://gamma.umd.edu/researchdirections/autonomousdriving/weathersafe/) 155 | * [Test-Time Personalization with a Transformer for Human Pose Estimation](https://proceedings.neurips.cc/paper/2021/file/1517c8664be296f0d87d9e5fc54fdd60-Paper.pdf) NeurIPS'21 [[Code]](https://github.com/harry11162/TTP) 156 | * [Towards Robust and Adaptive Motion Forecasting: A Causal Representation Perspective](https://arxiv.org/abs/2111.14820) CVPR'22 [[Code]](https://github.com/vita-epfl/causalmotion) 157 | * [Ev-TTA: Test-Time Adaptation for Event-Based Object Recognition](https://arxiv.org/abs/2203.12247) CVPR'22 [[Code]](https://github.com/82magnolia/ev_tta) 158 | * [MM-TTA: Multi-Modal Test-Time Adaptation for 3D Semantic Segmentation](https://arxiv.org/abs/2204.12667) CVPR'22 [[Project]](https://www.nec-labs.com/~mas/MM-TTA/) 159 | * [Source-Free Object Detection by Learning to Overlook Domain Style](https://openaccess.thecvf.com/content/CVPR2022/html/Li_Source-Free_Object_Detection_by_Learning_To_Overlook_Domain_Style_CVPR_2022_paper.html) CVPR'22 [[Code]](https://github.com/Flashkong/Source-Free-Object-Detection-by-Learning-to-Overlook-Domain-Style) 160 | * [On the Road to Online Adaptation for Semantic Image Segmentation](https://openaccess.thecvf.com/content/CVPR2022/papers/Volpi_On_the_Road_to_Online_Adaptation_for_Semantic_Image_Segmentation_CVPR_2022_paper.pdf) CVPR'22 161 | * [Test-Time Training Can Close the Natural Distribution Shift Performance Gap in Deep Learning Based Compressed Sensing](https://proceedings.mlr.press/v162/darestani22a/darestani22a.pdf) ICML'22 [[Code]](https://github.com/MLI-lab/ttt_for_deep_learning_cs) 162 | * [The Dual Form of Neural Networks Revisited: Connecting Test Time Predictions to Training Patterns via Spotlights of Attention 163 | ](https://proceedings.mlr.press/v162/irie22a/irie22a.pdf) ICML'22 [[Code]](https://github.com/robertcsordas/linear_layer_as_attention) 164 | * [Source-free Video Domain Adaptation by Learning Temporal Consistency for Action Recognition](https://arxiv.org/abs/2203.04559) ECCV'22 [[Code]](https://github.com/xuyu0010/ATCoN) 165 | * [AuxAdapt: Stable and Efficient Test-Time Adaptation for Temporally Consistent Video Semantic Segmentation](https://openaccess.thecvf.com/content/WACV2022/papers/Zhang_AuxAdapt_Stable_and_Efficient_Test-Time_Adaptation_for_Temporally_Consistent_Video_WACV_2022_paper.pdf) WACV'22 166 | * [Test-Time Prompt Tuning for Zero-Shot Generalization in Vision-Language Models](https://openreview.net/forum?id=e8PVEkSa4Fq) NeurIPS'22 [[Project]](https://azshue.github.io/TPT/) 167 | * [TTA-COPE: Test-Time Adaptation for Category-Level Object Pose Estimation](https://arxiv.org/abs/2303.16730) CVPR'23 [[Project]](https://sites.google.com/view/taeyeop-lee/ttacope) 168 | * [SfM-TTR: Using Structure from Motion for Test-Time Refinement of Single-View Depth Networks](https://arxiv.org/abs/2211.13551) CVPR'23 [[Code]](https://github.com/serizba/SfM-TTR) 169 | * [Video Test-Time Adaptation for Action Recognition](https://arxiv.org/abs/2211.15393) CVPR'23 [[Project]](https://wlin-at.github.io/vitta) 170 | 171 | --------------------------------------------------------------------------------