CEDR

This repository is for Contrastive Embedding Distribution Refinement and Entropy-Aware Attention Network (CEDR) introduced in the following paper:

"Contrastive Embedding Distribution Refinement and Entropy-Aware Attention for 3D Point Cloud Classification"

Updates

• 03/01/2022 The paper is currently under review, and the codes will be released in the future.
• 06/01/2022 codes for both model.py and main.py are available now.
• 10/01/2022 Update a pre-trained model (OA: 82.90%, mAcc: 80.60%) on ScanObjectNN via google drive.
• 10/01/2022 Pre-trained model (OA: 93.10%, mAcc: 91.10%) on ModelNet40 is available at google drive.

Network Architecture

Implementation Platforms

• Python 3.6
• Pytorch 0.4.0 with Cuda 9.1
• Higher Python/Pytorch/Cuda versions should also be compatible

ModelNet40 Experiment

Test the pre-trained model:

• download ModelNet40, unzip and move modelnet40_ply_hdf5_2048 folder to ./data

• put the pre-trained model under ./checkpoints/modelnet

• then run (more settings can be modified in main.py):

python main.py --exp_name=gbnet_modelnet40_eval --model=gbnet --dataset=modelnet40 --eval=True --model_path=checkpoints/modelnet/gbnet_modelnet40.t7

ScanObjectNN Experiment

Test the pre-trained model:

• download ScanObjectNN, and extract both training_objectdataset_augmentedrot_scale75.h5 and test_objectdataset_augmentedrot_scale75.h5 files to ./data
• put the pre-trained model under ./checkpoints/gbnet_scanobjectnn
• then run (more settings can be modified in main.py):

python main.py --exp_name=gbnet_scanobjectnn_eval --model=gbnet --dataset=ScanObjectNN --eval=True --model_path=checkpoints/gbnet_scanobjectnn/gbnet_scanobjectnn.t7

Pre-trained Models

• Python 3.6, Pytorch 0.4.0, Cuda 9.1
• 8 GeForce RTX 2080Ti GPUs
• using default training settings as in main.py

Model	Dataset	#Points	Data Augmentation	Performance on Test Set	Download Link
PointNet++	ModelNet40	1024	random scaling and translation	overall accuracy: 93.1% average class accuracy: 91.1%	google drive
GBNet	ScanObjectNN	1024	random scaling and translation	overall accuracy: 82.9% average class accuracy: 80.6%	google drive

Acknowledgement

The code is built on GBNet. We thank the authors for sharing the codes. We also thank the Big Data Center of Southeast University for providing the facility support on the numerical calculations in this paper.