semantic-gaussians
[TCSVT 2026] Official implemetation of the paper "Semantic Gaussians: Open-Vocabulary Scene Understanding with 3D Gaussian Splatting".
File Explorer
Download Latest Version (.zip)- framework.png
- teaser.png
- distill_scannet.yaml
- eval.yaml
- fusion_mipnerf360.yaml
- fusion_mvimgnet.yaml
- fusion_panoptic.yaml
- fusion_scannet.yaml
- official_train.yaml
- view_scannet.yaml
- __init__.py
- label_mapping.py
- scannet_constants.py
- scannetv2-labels.modified.tsv
- scannetv2_test.txt
- scannetv2_train.txt
- scannetv2_val.txt
- __init__.py
- augmentation.py
- feature_dataset.py
- fusion_utils.py
- encoding_models.py
- models.py
- __init__.py
- evaluation.py
- logger.py
- utils.py
- vis.py
- architecture.png
- qualitative_results.png
- fold0.pkl
- fold1.pkl
- fold2.pkl
- fold3.pkl
- fold0.pkl
- fold1.pkl
- fold2.pkl
- fold3.pkl
- test.txt
- trn.txt
- val.txt
- fold0.txt
- fold1.txt
- fold2.txt
- fold3.txt
- fold0.txt
- fold1.txt
- fold2.txt
- fold3.txt
- coco.py
- dataset.py
- fss.py
- pascal.py
- conv4d.py
- correlation.py
- feature.py
- hsnet.py
- learner.py
- README.md
- test.py
- train.py
- cat1.jpeg
- ade20k_objectInfo150.txt
- fewshot_coco.txt
- fewshot_fss.txt
- fewshot_pascal.txt
- lseg_blocks.py
- lseg_blocks_zs.py
- lseg_net.py
- lseg_net_zs.py
- lseg_vit.py
- lseg_vit_zs.py
- lseg_module.py
- lseg_module_zs.py
- lsegmentation_module.py
- lsegmentation_module_zs.py
- lseg_blocks.py
- lseg_blocks_zs.py
- lseg_net.py
- lseg_net_zs.py
- lseg_vit.py
- lseg_vit_zs.py
- lseg_module.py
- lseg_module_zs.py
- lsegmentation_module.py
- lsegmentation_module_zs.py
- LICENSE
- lseg_app.py
- lseg_demo.ipynb
- prepare_ade20k.py
- README.MD
- setup.cfg
- setup.py
- test_lseg.py
- test_lseg_zs.py
- train_lseg.py
- utils.py
- __init__.py
- swintransformer.py
- text_encoder.py
- vlpart.py
- vlpart_fast_rcnn.py
- vlpart_roi_heads.py
- vocab.py
- __init__.py
- gaussian_model.py
- lseg_predictor.py
- mink_unet.py
- openseg_predictor.py
- render_utils.py
- renderer.py
- resnet_base.py
- samclip_predictor.py
- vlpart_predictor.py
- __init__.py
- blender_loader.py
- camera.py
- colmap_loader.py
- scannet_loader.py
- scene.py
- __init__.py
- auxiliary.h
- backward.cu
- backward.h
- config.h
- forward.cu
- forward.h
- rasterizer.h
- rasterizer_impl.cu
- rasterizer_impl.h
- glm
- stbi_image_write.h
- .gitignore
- CMakeLists.txt
- ext.cpp
- LICENSE.md
- rasterize_points.cu
- rasterize_points.h
- README.md
- setup.py
- auxiliary.h
- backward.cu
- backward.h
- config.h
- forward.cu
- forward.h
- rasterizer.h
- rasterizer_impl.cu
- rasterizer_impl.h
- __init__.py
- glm
- stbi_image_write.h
- CMakeLists.txt
- ext.cpp
- LICENSE.md
- rasterize_points.cu
- rasterize_points.h
- README.md
- setup.py
- __init__.py
- common.py
- image_encoder.py
- mask_decoder.py
- mask_decoder_hq.py
- prompt_encoder.py
- sam.py
- transformer.py
- __init__.py
- amg.py
- onnx.py
- transforms.py
- __init__.py
- automask.py
- automatic_mask_generator.py
- build_sam.py
- build_sam_hq.py
- predictor.py
- .flake8
- CODE_OF_CONDUCT.md
- CONTRIBUTING.md
- LICENSE
- linter.sh
- README.md
- setup.cfg
- setup.py
- .gitkeep
- ext.cpp
- setup.py
- simple_knn.cu
- simple_knn.h
- spatial.cu
- spatial.h
- scannet_sens_reader.py
- unzip_label_filt.py
- camera_utils.py
- dataset_utils.py
- general_utils.py
- graphics_utils.py
- loss_utils.py
- metric.py
- sh_utils.py
- system_utils.py
- .gitignore
- .gitmodules
- distill.py
- environment.yml
- eval_segmentation.py
- fusion.py
- LICENSE
- README.md
- requirements.txt
- train.py
- view_viser.py
# Installation Guide
1. Get the code
git clone https://github.com/sharinka0715/semantic-gaussians
Downloads the entire project code from GitHub to your computer.
cd semantic-gaussians
Moves into the project folder you just downloaded.
2. Official Install Script
Easy RecommendedPrerequisites
- Python 3 Python is required to use pip.
pip install git+https://github.com/NVIDIA/MinkowskiEngine -v --no-deps --install-option="--blas_include_dirs=${CONDA_PREFIX}/include" --install-option="--blas=openblas"
Installs the package published on PyPI directly β no need to clone the source.
After installing, open a new terminal and run the program's version command (e.g. --version) to confirm it worked.
Pulled directly from this repo's README.
3. CMake
MediumPrerequisites
β οΈ This is a large repository, so this method may point to an internal sub-package rather than the actual core product. Check the full README as well.
cd submodules/channel-rasterization
This project's files live in a subfolder, so move into it first.
mkdir build && cd build
Creates a folder to hold the build output and moves into it.
cmake ..
Analyzes the source code and generates build configuration files (must be run inside the build folder).
make
Compiles the code based on the generated build configuration to produce an executable.
Check that an executable was created inside the build folder, then run it directly (e.g. ./build/app_name).
4. Python
EasyPrerequisites
pip install -r requirements.txt
Installs the Python libraries listed in requirements.txt (or similar).
pip install git+https://github.com/NVIDIA/MinkowskiEngine -v --no-deps --install-option="--blas_include_dirs=${CONDA_PREFIX}/include" --install-option="--blas=openblas"
Installs the package published on PyPI directly β no need to clone the source.
If it runs without errors and prints output in the terminal, it worked.
Pulled directly from this repo's README.
// repository documentation
Was this content helpful?
(0 ratings)