This code is supplementary to the submitted paper with ID 50 Title: SimpleView++: Neighborhood Views for Point Cloud Classification Authors: Shivanand Venkanna Sheshappanavar and Chandra Kambhamettu

Install Libraries

We recommend you first install Anaconda and create a virtual environment.

conda create --name svpp python=3.7.5 -y conda activate svpp pip install -r requirements.txt conda install sed -y export TORCH_CUDA_ARCH_LIST="6.0;6.1;6.2;7.0;7.5;8.6" cd pointnet2_pyt && pip install -e . && cd ..

Download Datasets and Pre-trained Models

cd data wget https://shapenet.cs.stanford.edu/media/modelnet40_ply_hdf5_2048.zip --no-check-certificate unzip modelnet40_ply_hdf5_2048.zip cd ..

Running Experiments

Training

python main.py --exp-config configs/dgcnn_simpleview_run_1.yaml

Testing

python main.py --entry test --exp-config configs/dgcnn_simpleview_run_1.yaml --model-path runs2/dgcnn_simpleview_run_1/model_best_test.pth

Note: this is a basic configuration set at 32x32 resolutions: edit line 4 of mv_utils.py and line 50 of mv.py in models directory to increase the resolutions We will provide detailed steps to recreate all results upon acceptance (we will also release the pre-trained models)

Acknowlegements

We would like to thank the authors of the following repositories for sharing their code.

• SimpleView: Revisiting Point Cloud Shape Classification with a Simple and Effective Baseline: 1
• PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation: 1, 2
• PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space: 1, 2
• Relation-Shape Convolutional Neural Network for Point Cloud Analysis: 1
• Dynamic Graph CNN for Learning on Point Clouds: 1