PointACL:Adversarial Contrastive Learning for Robust Point Clouds Representation under Adversarial Attack(ICASSP 2023)

This is the official code implementation for the paper "PointACL:Adversarial Contrastive Learning for Robust Point Clouds Representation under Adversarial Attack"

Downstream Tasks

• Shape Classification
• Semantic Segmentation

Datasets

Please download the used dataset with the following links:

• ModelNet40: https://shapenet.cs.stanford.edu/media/modelnet40_normal_resampled.zip

Pre-training

SimCLR framework

Please run the following command:

python BYOL/train.py

You need to edit the config file BYOL/config/config.yaml to switch different backbone architectures

Evaluation on Downstream Tasks

Coming soon