Compositionally Generalizable 3D Structure Prediction

In this work, We bring in the concept of compositional generalizability and factorizes the 3D shape reconstruction problem into proper sub-problems, each of which is tackled by a carefully designed neural sub-module with generalizability guarantee. Experiments on PartNet show that we achieve superior performance than baseline methods, which validates our problem factorization and network designs. Link to our paper.

Check our YouTube videos below for more details.

If you find this project useful for your research, please cite:

@article{han2020compositionally,
author = {Han, Songfang and Gu, Jiayuan and Mo, Kaichun and Yi, Li and Hu, Siyu and Chen, Xuejin and Su, Hao},
title = {{C}ompositionally {G}eneralizable 3{D} {S}tructure {P}rediction},
journal = {arXiv preprint},
year = {2020}}

How to use

Installation

• Check out the source code

  git clone https://github.com/hansongfang/CompNet.git && cd CompNet

• Install dependencies

  conda env create -f environment.yml && conda activate CompNet

• Compile CUDA extensions

  cd common_3d && bash compile.sh

Training and evaluating

Follow instructions in CompNet README

License

MIT Licence

Updates

• [Sep 16, 2021] Preliminary version of Data and Code released.