This is a framework for running common deep learning models for point cloud analysis tasks against classic benchmark. It heavily relies on pytorch geometric and hydra core.

Documentation | Pytorch Geometric | Facebook Hydra

The framework allows lean and yet complex model to be built with minimum effort and great reproducibility.

COMPACT API

  # PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space (https://arxiv.org/abs/1706.02413)
  pointnet2:
      type: pointnet2
      down_conv:
          module_name: SAModule
          ratios: [0.2, 0.25]
          radius: [0.2, 0.4]
          down_conv_nn: [[FEAT + 3, 64, 64, 128], [128 + 3, 128, 128, 256]]
      up_conv:
          module_name: FPModule
          up_conv_nn: [[1024 + 256, 256, 256], [256 + 128, 256, 128], [128 + FEAT, 128, 128, 128]]
          up_k: [1, 3, 3]   
          skip: True    
      innermost:
          module_name: GlobalBaseModule
          aggr: max
          nn: [256 + 3, 256, 512, 1024]
      mlp_cls: 
          nn: [128, 128, 128, 128, 128]
          dropout: 0.5

Getting started

You will first need to install poetry in order to setup a virtual environments and install the relevant packages, then run

poetry install

This will install all required dependencies in a new virtual environment.

Train pointnet++ on Segmentation task for dataset shapenet

poetry run python train.py experiment.name=pointnet2 experiment.data=shapenet

And you should see something like that

Benchmark

S3DIS

RSConv 2LD

Contributing

We use autopep8 for formating with the following options: --max-line-length 120 --ignore E402,E226,E24,W50,W690

Contributers

• Thomas Chaton
• Nicolas Chaulet
• Tristan Heywood