《A-CNN: Annularly Convolutional Neural Networks on Point Clouds》(2019)

Related tags

Deep Learning a-cnn
Overview

A-CNN: Annularly Convolutional Neural Networks on Point Clouds

Created by Artem Komarichev, Zichun Zhong, Jing Hua from Department of Computer Science, Wayne State University.

teaser image

Introduction

Our paper (arXiV) proposes a new approach to define and compute convolution directly on 3D point clouds by the proposed annular convolution.

To appear, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2019

A-CNN usage

We provide the code of A-CNN model that was tested with Tensorflow 1.3.0, CUDA 8.0, and python 3.6 on Ubuntu 16.04. We run all our experiments on a single NVIDIA Titan Xp GPU with 12GB GDDR5X.

  • Classification Task

    Download ModelNet-40 dataset first. Point clouds are sampled from meshes with 10K points (XYZ + normals) per shape and provided by PointNet++.

    To train a classification A-CNN model on ModelNet-40 dataset type the following command:

      python train.py

    To evaluate a trained model run the following script:

      python evaluate.py

  • Part Segmentation Task

    Download ShapeNet-part dataset first. Each point cloud represented by 2K points (XYZ + normals) and provided by PointNet++.

    To train a part segmentation A-CNN model on ShapeNet-part dataset type the following commands:

      cd part_segm
  python train.py

    To evaluate a trained segmentation model run the following script:

      ./evaluate_job.sh

  • Semantic Segmentation Task

    Download S3DIS and ScanNet datasets provided by PointNet/PointNet++. S3DIS contains XYZ + RGB information. ScanNet only has geometry information (XYZ only), no color.

    To estimate normals we used PCL library. The script to estimate normals for ScanNet data could be found here:

      cd scannet/normal_estimation
  ./run.sh

Citation

If you find our work useful in your research, please cite our work:

@InProceedings{komarichev2019acnn,
    title={A-CNN: Annularly Convolutional Neural Networks on Point Clouds},
    author={Komarichev, Artem and Zhong, Zichun and Hua, Jing},
    booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
    year={2019}
}
Comments
  • performance on shapenet

    performance on shapenet

    @artemkomarichev Hi, thanks for your code. I trained A-CNN on ShapeNet-Part, and I got 85.5 overall accuracy (86.1 reported in paper). I don't know why.

    eval mIoU of Airplane: 0.830114 eval mIoU of Bag: 0.805713 eval mIoU of Cap: 0.856973 eval mIoU of Car: 0.793102 eval mIoU of Chair: 0.911208 eval mIoU of Earphone: 0.769749 eval mIoU of Guitar: 0.911113 eval mIoU of Knife: 0.845159 eval mIoU of Lamp: 0.840812 eval mIoU of Laptop: 0.960384 eval mIoU of Motorbike: 0.725803 eval mIoU of Mug: 0.953242 eval mIoU of Pistol: 0.825036 eval mIoU of Rocket: 0.637810 eval mIoU of Skateboard: 0.761378 eval mIoU of Table: 0.829875 eval mean mIoU: 0.828592 eval mean mIoU (all shapes): 0.855444

    opened by inshallahzz 1
  • Low accuracy

    Low accuracy

    Thank you for your outstanding work. I found some problems when using your code. After training for 200 epochs, the highest accuracy rate is less than 50%. Is there a way to solve this problem?

    opened by ke-dev 0
  • download dateset

    download dateset

    Hello! I am very interested in your research. But I don’t know where to download the dataset. The link you gave has been deleted！

    Can you tell me where to download the dataset？

    I look forward to your response. Thanks a lot.

    opened by yy-zhang832 0
  • without normals experiment

    without normals experiment

    Hello! I'm very interested in your research. I want to run your classification code in other dataset, which is only providing location. And I wonder that how you order points without normals, there is only code for with normals.

    I'm look forward to your reply and thank you very much.

    opened by tkdguraa 1
  • code for semantic segmentation

    code for semantic segmentation

    Hello ! Thanks for sharing code.Your work is very interesting. You have given code for both classification and part segmentation,but there is no code for semantic segmentation. can you please provide it? Thanks Look forward to hear from you.

    opened by Emenent758 1
  • versions of modelnet dataset

    versions of modelnet dataset

    Hi ! You used the following version of modelnet dataset (it has 10 k points per shape) https://shapenet.cs.stanford.edu/media/modelnet40_normal_resampled.zip

    while pointnet2 has used an other version which has 2048 points per shape. https://shapenet.cs.stanford.edu/media/modelnet40_ply_hdf5_2048.zip)

    If versions of datasets are different then is the comparison of accuracy produced by pointnet++ and your network justified?

    I look forward to your response. Thanks a lot.

    opened by rabbiahassan 1
Owner
Artёm Komarichev
Artёm Komarichev
GitHub
Code for "PV-RAFT: Point-Voxel Correlation Fields for Scene Flow Estimation of Point Clouds", CVPR 2021

PV-RAFT This repository contains the PyTorch implementation for paper "PV-RAFT: Point-Voxel Correlation Fields for Scene Flow Estimation of Point Clou

Yi Wei 43 Dec 5, 2022
Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Clouds (CVPR 2022, Oral)

Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Clouds (CVPR 2022, Oral) This is the official implementat

Yifan Zhang 259 Dec 25, 2022
This project is the official implementation of our accepted ICLR 2021 paper BiPointNet: Binary Neural Network for Point Clouds.

BiPointNet: Binary Neural Network for Point Clouds Created by Haotong Qin, Zhongang Cai, Mingyuan Zhang, Yifu Ding, Haiyu Zhao, Shuai Yi, Xianglong Li

Haotong Qin 59 Dec 17, 2022
This is the code repository implementing the paper "TreePartNet: Neural Decomposition of Point Clouds for 3D Tree Reconstruction".

TreePartNet This is the code repository implementing the paper "TreePartNet: Neural Decomposition of Point Clouds for 3D Tree Reconstruction". Depende

刘彦超 34 Nov 30, 2022
Neural-Pull: Learning Signed Distance Functions from Point Clouds by Learning to Pull Space onto Surfaces(ICML 2021)

Neural-Pull: Learning Signed Distance Functions from Point Clouds by Learning to Pull Space onto Surfaces(ICML 2021) This repository contains the code

null 149 Dec 15, 2022
This repository implements and evaluates convolutional networks on the Möbius strip as toy model instantiations of Coordinate Independent Convolutional Networks.

Orientation independent Möbius CNNs This repository implements and evaluates convolutional networks on the Möbius strip as toy model instantiations of

Maurice Weiler 59 Dec 9, 2022
A PyTorch implementation of "Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks" (KDD 2019).

ClusterGCN ⠀⠀ A PyTorch implementation of "Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks" (KDD 2019). A

Benedek Rozemberczki 697 Dec 27, 2022
Scripts for training an AI to play the endless runner Subway Surfers using a supervised machine learning approach by imitation and a convolutional neural network (CNN) for image classification

About subwAI subwAI - a project for training an AI to play the endless runner Subway Surfers using a supervised machine learning approach by imitation

null 82 Jan 1, 2023
NFT-Price-Prediction-CNN - Using visual feature extraction, prices of NFTs are predicted via CNN (Alexnet and Resnet) architectures.

NFT-Price-Prediction-CNN - Using visual feature extraction, prices of NFTs are predicted via CNN (Alexnet and Resnet) architectures.

null 5 Nov 3, 2022
(CVPR 2021) PAConv: Position Adaptive Convolution with Dynamic Kernel Assembling on Point Clouds

PAConv: Position Adaptive Convolution with Dynamic Kernel Assembling on Point Clouds by Mutian Xu*, Runyu Ding*, Hengshuang Zhao, and Xiaojuan Qi. Int

CVMI Lab 228 Dec 25, 2022
Official PyTorch implementation of CAPTRA: CAtegory-level Pose Tracking for Rigid and Articulated Objects from Point Clouds

CAPTRA: CAtegory-level Pose Tracking for Rigid and Articulated Objects from Point Clouds Introduction This is the official PyTorch implementation of o

Yijia Weng 96 Dec 7, 2022
(CVPR 2021) Back-tracing Representative Points for Voting-based 3D Object Detection in Point Clouds

BRNet Introduction This is a release of the code of our paper Back-tracing Representative Points for Voting-based 3D Object Detection in Point Clouds,

null 86 Oct 5, 2022
Self-Supervised Learning for Domain Adaptation on Point-Clouds

Self-Supervised Learning for Domain Adaptation on Point-Clouds Introduction Self-supervised learning (SSL) allows to learn useful representations from

Idan Achituve 66 Dec 20, 2022
Rendering Point Clouds with Compute Shaders

Compute Shader Based Point Cloud Rendering This repository contains the source code to our techreport: Rendering Point Clouds with Compute Shaders and

Markus Schütz 460 Jan 5, 2023
This is a package for LiDARTag, described in paper: LiDARTag: A Real-Time Fiducial Tag System for Point Clouds

LiDARTag Overview This is a package for LiDARTag, described in paper: LiDARTag: A Real-Time Fiducial Tag System for Point Clouds (PDF)(arXiv). This wo

University of Michigan Dynamic Legged Locomotion Robotics Lab 159 Dec 21, 2022
Uncertainty-aware Semantic Segmentation of LiDAR Point Clouds for Autonomous Driving

SalsaNext: Fast, Uncertainty-aware Semantic Segmentation of LiDAR Point Clouds for Autonomous Driving Abstract In this paper, we introduce SalsaNext f

null 308 Jan 4, 2023
Code for "CloudAAE: Learning 6D Object Pose Regression with On-line Data Synthesis on Point Clouds" @ICRA2021

CloudAAE This is an tensorflow implementation of "CloudAAE: Learning 6D Object Pose Regression with On-line Data Synthesis on Point Clouds" Files log:

Gee 35 Nov 14, 2022
This repo is a PyTorch implementation for Paper "Unsupervised Learning for Cuboid Shape Abstraction via Joint Segmentation from Point Clouds"

Unsupervised Learning for Cuboid Shape Abstraction via Joint Segmentation from Point Clouds This repository is a PyTorch implementation for paper: Uns

Kaizhi Yang 42 Dec 9, 2022
Unsupervised 3D Human Mesh Recovery from Noisy Point Clouds

Unsupervised 3D Human Mesh Recovery from Noisy Point Clouds Xinxin Zuo, Sen Wang, Minglun Gong, Li Cheng Prerequisites We have tested the code on Ubun

null 41 Dec 12, 2022