Enhancing 3D Semantic Scene Completion with Refinement Module (ESSC-RM)

July 14, 2025 · View on GitHub

Network Architecture

CGFormer is developed based on the official OccFormer codebase and the installation follows similar steps.

a. Create a conda virtual environment and activate

conda create -n ESSC python=3.8 -y
conda activate ESSC

b. Install PyTorch and torchvision following the official instructions

conda install pytorch==2.0.0 torchvision==0.15.0 torchaudio==2.0.0 pytorch-cuda=11.7 -c pytorch -c nvidia

c. Install other dependencies, like timm, einops, torchmetrics, spconv, pytorch-lightning, etc.

cd docs
pip install -r requirements.txt

d. Install Natten

pip3 install natten==0.17.4+torch200cu117 -f https://shi-labs.com/natten/wheels

e. Install OpenMMLab ToolKit

pip install mmengine==0.10.6
pip install mmcv==2.1.0 -f https://download.openmmlab.com/mmcv/dist/cu117/torch2.0/index.html
pip install mmdet==3.3.0

f. Install environment

pip install -e /u/home/caoh/projects/MA_Jiachen/3DPNA
pip install -e .

g. Fix bugs (known now)

Please check VLGSSC