This repository contains code for the CVNet framework as described in our CVPR 2022 paper.

• Python 3 (we used v3.6.5)
• PyTorch (we used v1.2.0)
• PIL
• scipy and associated packages
• tqdm
• ChamferDistance

1. Download datasets from here
2. Unzip datasets into your desired directory datasets
3. Modify these directories in package/config/config.ini, and the directory where you intend to keep results
4. Complete train/val/test split by split_datasets.py, our divided txt files are placed in package/datasets
5. Run experiments with runner.py
6. Download models from here

1. Darnet (https://github.com/dcheng-utoronto/darnet)
2. ChamferDistance (https://github.com/ThibaultGROUEIX/ChamferDistancePytorch)

Please contact me at xuziqiang@njust.edu.cn for questions.