The code for CVPR 2025 paper "Reconstructing Close Human Interaction with Appearance and Proxemics Reasoning"

Buzhen Huang, Chen Li, Chongyang Xu, Dongyue Lu, Jinnan Chen, Yangang Wang, Gim Hee Lee
[Project] [Paper] [Dataset]

Installation

The code is tested on Ubuntu 22.04 with a single RTX 3090 GPU.

conda create -n closeapp python=3.10
pip install torch==2.3.0 torchvision==0.18.0 torchaudio==2.3.0 --index-url https://download.pytorch.org/whl/cu118

Then, compile diff-gaussian-rasterization as in 3DGS repository.

SDF loss and BVH_CUDA are required for evaluating penetration.

Download the official SMPL model from SMPL and SMPLify website and place them in data/smpl/smpl.

Download data from Baidu Netdisk or Google Drive.

Demo

python train.py -s data/preprocess_data/04305 -m output/04305 --train_stage=1 --save_render --use_appearance --save_params

Data processing

To process your own images, please refer to the Data Processing Tutorial.

Citation

If you find this code or dataset useful for your research, please consider citing the paper.

@inproceedings{huang2025reconstructing,
  title={Reconstructing Close Human Interaction with Appearance and Proxemics Reasoning},
  author={Huang, Buzhen and Li, Chen and Xu, Chongyang and Lu, Dongyue and Chen, Jinnan and Wang, Yangang and Lee, Gim Hee},
  booktitle={Proceedings of the Computer Vision and Pattern Recognition Conference},
  pages={17475--17485},
  year={2025}
}
@inproceedings{huang2024closely,
  title={Closely interactive human reconstruction with proxemics and physics-guided adaption},
  author={Huang, Buzhen and Li, Chen and Xu, Chongyang and Pan, Liang and Wang, Yangang and Lee, Gim Hee},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={1011--1021},
  year={2024}
}

Acknowledgments

Some of the code are based on the following works.
CloseInt
GaussianAvatar
aitviewer