├── .gitignore
├── README.md
└── media
    ├── 7scenes_office.mp4
    ├── 7scenes_office.png
    ├── 7scenes_redkitchen.mp4
    ├── 7scenes_redkitchen.png
    ├── architecture.png
    ├── bf_apt1.mp4
    ├── bf_apt1.png
    ├── bf_office0.mp4
    ├── bf_office0.png
    ├── o3d_eg.gif
    ├── rerun_eg.gif
    ├── scannet_0000.mp4
    ├── scannet_0000.png
    ├── scannet_0054.mp4
    ├── scannet_0054.png
    ├── teaser.gif
    ├── teaser.png
    ├── tumrgbd_floor.mp4
    ├── tumrgbd_floor.png
    ├── tumrgbd_room.mp4
    ├── tumrgbd_room.png
    ├── vista_logo.png
    └── vista_slam.mp4


/.gitignore:
--------------------------------------------------------------------------------
  1 | data
  2 | checkpoints
  3 | results
  4 | stuffs
  5 | recon
  6 | visualization
  7 | replica_gt
  8 | tmp
  9 | 
 10 | # Byte-compiled / optimized / DLL files
 11 | __pycache__/
 12 | *.py[cod]
 13 | *$py.class
 14 | 
 15 | # C extensions
 16 | *.so
 17 | 
 18 | # Distribution / packaging
 19 | .Python
 20 | build/
 21 | develop-eggs/
 22 | dist/
 23 | downloads/
 24 | eggs/
 25 | .eggs/
 26 | lib/
 27 | lib64/
 28 | parts/
 29 | sdist/
 30 | var/
 31 | wheels/
 32 | pip-wheel-metadata/
 33 | share/python-wheels/
 34 | *.egg-info/
 35 | .installed.cfg
 36 | *.egg
 37 | MANIFEST
 38 | 
 39 | # PyInstaller
 40 | #  Usually these files are written by a python script from a template
 41 | #  before PyInstaller builds the exe, so as to inject date/other infos into it.
 42 | *.manifest
 43 | *.spec
 44 | 
 45 | # Installer logs
 46 | pip-log.txt
 47 | pip-delete-this-directory.txt
 48 | 
 49 | # Unit test / coverage reports
 50 | htmlcov/
 51 | .tox/
 52 | .nox/
 53 | .coverage
 54 | .coverage.*
 55 | .cache
 56 | nosetests.xml
 57 | coverage.xml
 58 | *.cover
 59 | *.py,cover
 60 | .hypothesis/
 61 | .pytest_cache/
 62 | 
 63 | # Translations
 64 | *.mo
 65 | *.pot
 66 | 
 67 | # Django stuff:
 68 | *.log
 69 | local_settings.py
 70 | db.sqlite3
 71 | db.sqlite3-journal
 72 | 
 73 | # Flask stuff:
 74 | instance/
 75 | .webassets-cache
 76 | 
 77 | # Scrapy stuff:
 78 | .scrapy
 79 | 
 80 | # Sphinx documentation
 81 | docs/_build/
 82 | 
 83 | # PyBuilder
 84 | target/
 85 | 
 86 | # Jupyter Notebook
 87 | .ipynb_checkpoints
 88 | 
 89 | # IPython
 90 | profile_default/
 91 | ipython_config.py
 92 | 
 93 | # pyenv
 94 | .python-version
 95 | 
 96 | # pipenv
 97 | #   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
 98 | #   However, in case of collaboration, if having platform-specific dependencies or dependencies
 99 | #   having no cross-platform support, pipenv may install dependencies that don't work, or not
100 | #   install all needed dependencies.
101 | #Pipfile.lock
102 | 
103 | # PEP 582; used by e.g. github.com/David-OConnor/pyflow
104 | __pypackages__/
105 | 
106 | # Celery stuff
107 | celerybeat-schedule
108 | celerybeat.pid
109 | 
110 | # SageMath parsed files
111 | *.sage.py
112 | 
113 | # Environments
114 | .env
115 | .venv
116 | env/
117 | venv/
118 | ENV/
119 | env.bak/
120 | venv.bak/
121 | 
122 | # Spyder project settings
123 | .spyderproject
124 | .spyproject
125 | 
126 | # Rope project settings
127 | .ropeproject
128 | 
129 | # mkdocs documentation
130 | /site
131 | 
132 | # mypy
133 | .mypy_cache/
134 | .dmypy.json
135 | dmypy.json
136 | 
137 | # Pyre type checker
138 | .pyre/
139 | 
140 | ORBvoc.txt
141 | frontend_sta_weights.pth
142 | 
143 | *.code-workspace
144 | 
145 | preprocessing/slurm_scripts
146 | slam3r/slam/test
147 | test
148 | 
149 | exp_figures
150 | test*.py
151 | 
152 | preprocessing
153 | pretrains
154 | output


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
  1 | 
  2 | ### The full codebase will be released soon ...
  3 | ---
  4 | 
  5 | 
  6 | <!-- PROJECT LOGO -->
  7 | <h1 align="center" style="display: flex; align-items: center; justify-content: center; flex-wrap: wrap;">
  8 |   <img src="./media/vista_logo.png" alt="logo" width="60" style="margin-right: 10px;">
  9 |   ViSTA-SLAM: Visual SLAM with Symmetric Two-view Association
 10 | </h1>
 11 | 
 12 | 
 13 |   <p align="center">
 14 |     <a href="https://ganlinzhang.xyz" target="_blank"><strong>Ganlin Zhang<sup>1,2</sup></strong></a>
 15 |     ·
 16 |     <a href="https://enhanqian.github.io/" target="_blank"><strong>Shenhan Qian<sup>1,2</sup></strong></a>
 17 |     ·
 18 |     <a href="https://xiwang1212.github.io/homepage/" target="_blank"><strong>Xi Wang<sup>1,2,3</sup></strong></a>
 19 |     ·
 20 |     <a href="https://cvg.cit.tum.de/members/cremers" target="_blank"><strong>Daniel Cremers<sup>1,2</sup></strong></a>
 21 |   </p>
 22 |   <p align="center">
 23 |       <strong><sup>1 </sup>TU Munich,  <sup>2 </sup>MCML, <sup>3 </sup>ETH Zurich</strong>
 24 |       <strong><h4 align="center"><a href="https://arxiv.org/abs/2509.01584" target="_blank">Paper</a> | <a href="https://ganlinzhang.xyz/vista-slam/" target="_blank">Project Website</a></h4></strong>
 25 |   </strong></p>
 26 | 
 27 | <p align="center">
 28 |     <img src="./media/teaser.gif" alt="teaser_img" width="100%">
 29 |     <strong>ViSTA-SLAM</strong> is a real-time monocular dense SLAM pipeline that combines a Symmetric Two-view Association (STA) frontend with Sim(3) pose graph optimization and loop closure, enabling accurate camera trajectories and high-quality 3D scene reconstruction from RGB inputs.
 30 | </p>
 31 | 
 32 | ### Online visualization
 33 | <p align="center">
 34 |     <img src="./media/rerun_eg.gif" alt="rerun_eg" width="100%">
 35 | </p>
 36 | 
 37 | ### Pose graph and final results visualization
 38 | <p align="center">
 39 |     <img src="./media/o3d_eg.gif" alt="rerun_eg" width="60%">
 40 | </p>
 41 | 
 42 | <!-- <details open="open" style='padding: 10px; border-radius:5px 30px 30px 5px; border-style: solid; border-width: 1px;'>
 43 |   <summary>Table of Contents</summary>
 44 |   <ol>
 45 |     <li>
 46 |       <a href="#installation">Installation</a>
 47 |     </li>
 48 |     <li>
 49 |       <a href="#run">Run</a>
 50 |     </li>
 51 |     <li>
 52 |       <a href="#visualization">Visualization</a>
 53 |     </li>
 54 |     <li>
 55 |       <a href="#evaluation">Evaluation</a>
 56 |     </li>
 57 |     <li>
 58 |       <a href="#acknowledgement">Acknowledgement</a>
 59 |     </li>
 60 |     <li>
 61 |       <a href="#citation">Citation and Contact</a>
 62 |     </li>
 63 |   </ol>
 64 | </details>
 65 | 
 66 | 
 67 | 
 68 | 
 69 | ## Installation
 70 | 1. Clone the repo
 71 | ```bash
 72 | git clone https://github.com/zhangganlin/vista-slam.git
 73 | cd vista-slam
 74 | ```
 75 | 2. Install dependency
 76 | ```bash
 77 | # used by Bag of Words loop detection
 78 | sudo apt-get install libopencv-dev
 79 | ```
 80 | 
 81 | 2. Creating a new conda environment and install python dependencies.
 82 | ```bash
 83 | conda create -n vista python=3.11 cmake=3.31.2 -c conda-forge
 84 | conda activate vista
 85 | 
 86 | # install torch according to your cuda version
 87 | pip install torch==2.5.1 torchvision==0.20.1 torchaudio==2.5.1 xformers --index-url https://download.pytorch.org/whl/cu121
 88 | 
 89 | # install python binding of DBoW3, for loop detection 
 90 | cd DBoW3Py
 91 | pip install --no-build-isolation .
 92 | cd ..
 93 | 
 94 | # install other python dependencies
 95 | pip install -r requirements.txt
 96 | 
 97 | # optional: accelerate with CUDA-based RoPE
 98 | cd vista_slam/sta_model/pos_embed/curope
 99 | python setup.py build_ext --inplace
100 | cd ../../../../
101 | ```
102 | 
103 | 4. Download pretrained model.
104 | Download the pretained models from HuggingFace (https://huggingface.co/zhangganlin/vista_slam/tree/main), and put them inside the `pretrains` folder.
105 | * [frontend_sta_weights.pth](https://huggingface.co/zhangganlin/vista_slam/resolve/main/frontend_sta_weights.pth?download=true)
106 | * [ORBvoc.txt](https://huggingface.co/zhangganlin/vista_slam/resolve/main/ORBvoc.txt?download=true)
107 | 
108 | <details>
109 |   <summary>[Directory structure of pretrianed (click to expand)]</summary>
110 |   
111 | ```bash
112 |   .
113 |   └── pretrains
114 |         ├── frontend_sta_weights.pth
115 |         ├── ORBvoc.txt
116 |         └── README.md
117 | 
118 | ```
119 | </details>
120 | 
121 | ## Data preparation
122 | ```bash
123 | # TUM-RGBD
124 | bash scripts/download_tumrgbd.sh
125 | 
126 | # 7-Scenes
127 | bash scripts/download_7scenes.sh
128 | ```
129 | The datasets will be downloaded to `./datasets` by default, please change the downloading scripts if you prefer other paths.
130 | 
131 | 
132 | ## Run
133 | ### Quick start
134 | ```bash
135 | # Run 7-Scenes redkitchen
136 | python run.py --config configs/7scenes.yaml --images "datasets/7scenes/redkitchen/seq-01/*.color.png" --output output/redkitchen
137 | 
138 | # Run TUM-RGBD floor
139 | python run.py --config configs/tumrgbd.yaml --images "datasets/rgbd_dataset_freiburg1_floor/rgb/*.png" --output output/floor
140 | ```
141 | These commands will run ViSTA-SLAM on 7-Scenes `redkitchen` scene and TUM-RGBD `floor` scene, and output to `output/redkitchen` and `output/floor`.
142 | 
143 | ### Test on other sequential data
144 | ```bash
145 | python run.py --config configs/default.yaml --images "PATH/TO/IMAGES*.png/jpg" --output OUTPUT_FOLDER
146 | ```
147 | All adjustable configuration parameters are defined in configs/default.yaml, where explanations are also provided. You can modify them to suit your setup.
148 | 
149 | ## Visualization
150 | ### Online visualization
151 | ViSTA-SLAM provide online visualization via [Rerun](https://rerun.io/). You can either add `--vis` to the command line directly or adjust `rerun_vis: True` in the config file.
152 | 
153 | To visualize the result online, open Rerun client in one terminal,
154 | ```bash
155 | rerun
156 | ```
157 | and run ViSTA-SLAM in another terminal.
158 | ```
159 | python run.py --config configs/default.yaml --images "PATH/TO/IMAGES*.png/jpg" --output OUTPUT_FOLDER --vis
160 | ```
161 | <p align="center">
162 |     <img src="./media/rerun_eg.gif" alt="rerun_eg" width="100%">
163 | </p>
164 | 
165 | You can also run ViSTA-SLAM in a remote machine (i.e. cluster) and visualize it in the local machine, just adjust the `rerun_url` in the config file, replace `rerun+http://127.0.0.1:9876/proxy` with your local machine's ip with similar format. And open Rerun cilent in the local machine, run ViSTA-SLAM in the remote machine, the visualization will be shown automatically in the local Rerun client.
166 | 
167 | 
168 | ### Visualize final results
169 | We also provide a script to visualize the final results -- trajectory, reconstruction, and pose graph -- using Open3D. After running ViSTA-SLAM, run
170 | ```bash
171 | python scripts/vis_slam_results.py OUTPUT_FOLDER
172 | ```
173 | <p align="center">
174 |     <img src="./media/o3d_eg.gif" alt="rerun_eg" width="60%">
175 | </p>
176 | Here, the light blue frustums represent the camera poses, blue lines indicate edges between neighboring views, and orange lines correspond to loop closure edges.
177 | 
178 | ## Evaluation
179 | ViSTA-SLAM is mainly evaulated in TUM-RGBD and 7-Scenes, here we also provide the evaluation scripts.
180 | ```bash
181 | # For 7-Scenes
182 | python evaluation_7scenes.py --dataset_folder "datasets/7scenes" --output output/7scenes
183 | 
184 | # For TUM-RGBD
185 | python evaluation_tumrgbd.py --dataset_folder "datasets/tumrgbd" --output output/tumrgbd
186 | ```
187 | > **Note:** There may be minor differences between the released codebase and the results reported in the paper due to code refactoring and hardware variations, but the overall results should be largely consistent.
188 | 
189 | 
190 | ## Acknowledgement
191 | Our codebase is partially based on [Spann3r](https://github.com/HengyiWang/spann3r), [SLAM3R](https://github.com/PKU-VCL-3DV/SLAM3R) and [VGGT-SLAM](https://github.com/MIT-SPARK/VGGT-SLAM), we thank the authors for making these codebases publicly available. Our work would not have been possible without your great efforts! -->
192 | 
193 | 
194 | ## Citation
195 | 
196 | If you find our code or paper useful, please cite
197 | ```bibtex
198 | @misc{zhang2025vistaslam,
199 |       title={{ViSTA-SLAM}: Visual {SLAM} with Symmetric Two-view Association}, 
200 |       author={Ganlin Zhang and Shenhan Qian and Xi Wang and Daniel Cremers},
201 |       year={2025},
202 |       eprint={2509.01584},
203 |       archivePrefix={arXiv},
204 |       primaryClass={cs.CV},
205 |       url={https://arxiv.org/abs/2509.01584}, 
206 | }
207 | ```
208 | ## Contact
209 | Contact [Ganlin Zhang](mailto:ganlin.zhang@tum.de) for questions, comments and reporting bugs.
210 | 


--------------------------------------------------------------------------------
/media/7scenes_office.mp4:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/7scenes_office.mp4


--------------------------------------------------------------------------------
/media/7scenes_office.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/7scenes_office.png


--------------------------------------------------------------------------------
/media/7scenes_redkitchen.mp4:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/7scenes_redkitchen.mp4


--------------------------------------------------------------------------------
/media/7scenes_redkitchen.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/7scenes_redkitchen.png


--------------------------------------------------------------------------------
/media/architecture.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/architecture.png


--------------------------------------------------------------------------------
/media/bf_apt1.mp4:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/bf_apt1.mp4


--------------------------------------------------------------------------------
/media/bf_apt1.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/bf_apt1.png


--------------------------------------------------------------------------------
/media/bf_office0.mp4:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/bf_office0.mp4


--------------------------------------------------------------------------------
/media/bf_office0.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/bf_office0.png


--------------------------------------------------------------------------------
/media/o3d_eg.gif:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/o3d_eg.gif


--------------------------------------------------------------------------------
/media/rerun_eg.gif:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/rerun_eg.gif


--------------------------------------------------------------------------------
/media/scannet_0000.mp4:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/scannet_0000.mp4


--------------------------------------------------------------------------------
/media/scannet_0000.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/scannet_0000.png


--------------------------------------------------------------------------------
/media/scannet_0054.mp4:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/scannet_0054.mp4


--------------------------------------------------------------------------------
/media/scannet_0054.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/scannet_0054.png


--------------------------------------------------------------------------------
/media/teaser.gif:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/teaser.gif


--------------------------------------------------------------------------------
/media/teaser.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/teaser.png


--------------------------------------------------------------------------------
/media/tumrgbd_floor.mp4:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/tumrgbd_floor.mp4


--------------------------------------------------------------------------------
/media/tumrgbd_floor.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/tumrgbd_floor.png


--------------------------------------------------------------------------------
/media/tumrgbd_room.mp4:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/tumrgbd_room.mp4


--------------------------------------------------------------------------------
/media/tumrgbd_room.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/tumrgbd_room.png


--------------------------------------------------------------------------------
/media/vista_logo.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/vista_logo.png


--------------------------------------------------------------------------------
/media/vista_slam.mp4:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/zhangganlin/vista-slam/7d43ca22218f9d80053b81cd5516bd1a75573a3e/media/vista_slam.mp4


--------------------------------------------------------------------------------