All Projects → wentaoyuan → Pcn

wentaoyuan / Pcn

Licence: mit
Code for PCN: Point Completion Network in 3DV'18 (Oral)

Programming Languages

python
139335 projects - #7 most used programming language

Projects that are alternatives of or similar to Pcn

Point2Mesh
Meshing Point Clouds with Predicted Intrinsic-Extrinsic Ratio Guidance (ECCV2020)
Stars: ✭ 61 (-74.37%)
Mutual labels:  point-cloud, 3d-reconstruction
pyRANSAC-3D
A python tool for fitting primitives 3D shapes in point clouds using RANSAC algorithm
Stars: ✭ 253 (+6.3%)
Mutual labels:  point-cloud, 3d-reconstruction
NeuralPull
Implementation of ICML'2021:Neural-Pull: Learning Signed Distance Functions from Point Clouds by Learning to Pull Space onto Surfaces
Stars: ✭ 149 (-37.39%)
Mutual labels:  point-cloud, 3d-reconstruction
Silhouette-Guided-3D
PyTorch Code of our WACV2020 paper: Silhouette Guided Point Cloud Reconstruction beyond Occlusion
Stars: ✭ 36 (-84.87%)
Mutual labels:  point-cloud, 3d-reconstruction
Awsome deep geometry learning
A list of resources about deep learning solutions on 3D shape processing
Stars: ✭ 105 (-55.88%)
Mutual labels:  point-cloud, 3d-reconstruction
Meshlab
The open source mesh processing system
Stars: ✭ 2,619 (+1000.42%)
Mutual labels:  point-cloud, 3d-reconstruction
softpool
SoftPoolNet: Shape Descriptor for Point Cloud Completion and Classification - ECCV 2020 oral
Stars: ✭ 62 (-73.95%)
Mutual labels:  point-cloud, 3d-reconstruction
3PU pytorch
pytorch implementation of >>Patch-base progressive 3D Point Set Upsampling<<
Stars: ✭ 61 (-74.37%)
Mutual labels:  point-cloud, 3d-reconstruction
3d Machine Learning
A resource repository for 3D machine learning
Stars: ✭ 7,405 (+3011.34%)
Mutual labels:  point-cloud, 3d-reconstruction
Openmvs
open Multi-View Stereo reconstruction library
Stars: ✭ 1,842 (+673.95%)
Mutual labels:  point-cloud, 3d-reconstruction
Msn Point Cloud Completion
Morphing and Sampling Network for Dense Point Cloud Completion (AAAI2020)
Stars: ✭ 196 (-17.65%)
Mutual labels:  point-cloud, 3d-reconstruction
Patchmatchstereo
PatchMatch,倾斜窗口经典,效果极佳,OpenMVS&Colmap稠密匹配算法。完整实现,代码规范,注释清晰,博客教学,欢迎star!
Stars: ✭ 219 (-7.98%)
Mutual labels:  3d-reconstruction
Liblas
C++ library and programs for reading and writing ASPRS LAS format with LiDAR data
Stars: ✭ 211 (-11.34%)
Mutual labels:  point-cloud
Computervisiondatasets
Stars: ✭ 207 (-13.03%)
Mutual labels:  3d-reconstruction
Graph Cnn In 3d Point Cloud Classification
Code for A GRAPH-CNN FOR 3D POINT CLOUD CLASSIFICATION (ICASSP 2018)
Stars: ✭ 206 (-13.45%)
Mutual labels:  point-cloud
Cupoch
Robotics with GPU computing
Stars: ✭ 225 (-5.46%)
Mutual labels:  point-cloud
Kitti Dataset
Visualising LIDAR data from KITTI dataset.
Stars: ✭ 217 (-8.82%)
Mutual labels:  point-cloud
Frustum Convnet
The PyTorch Implementation of F-ConvNet for 3D Object Detection
Stars: ✭ 203 (-14.71%)
Mutual labels:  point-cloud
Point Transformer Pytorch
Implementation of the Point Transformer layer, in Pytorch
Stars: ✭ 199 (-16.39%)
Mutual labels:  point-cloud
Horizonnet
Pytorch implementation of HorizonNet: Learning Room Layout with 1D Representation and Pano Stretch Data Augmentation.
Stars: ✭ 195 (-18.07%)
Mutual labels:  3d-reconstruction

PCN: Point Completion Network

[paper] [data] [website]

Introduction

PCN is a learning-based shape completion method which directly maps a partial point cloud to a dense, complete point cloud without any voxelization. It is based on our 3DV 2018 publication PCN: Point Completion Network. Please refer to our project website or read our paper for more details.

Citation

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

@inProceedings{yuan2018pcn,
  title     = {PCN: Point Completion Network},
  author    = {Yuan, Wentao and Khot, Tejas and Held, David and Mertz, Christoph and Hebert, Martial},
  booktitle = {3D Vision (3DV), 2018 International Conference on},
  year      = {2018}
}

Usage

1) Prerequisite

  1. Install dependencies via pip3 install -r requirments.txt.
  2. Follow this guide to install Open3D for point cloud I/O.
  3. Build point cloud distance ops by running make under pc_distance. Make sure the paths in makefile are correct.
  4. Download trained models from Google Drive.

This code is built using Tensorflow 1.12 with CUDA 9.0 and tested on Ubuntu 16.04 with Python 3.5.

2) Demo

Run python3 demo.py. Use --input_path option to switch between the input examples in demo_data.

3) ShapeNet Completion

  1. Download ShapeNet test data in the shapenet folder on Google Drive. Specifically, this experiment requires test, test_novel, test.list and test_novel.list.
  2. Run python3 test_shapenet.py. Use --model_type option to choose different model architectures. Type python3 test_shapenet.py -h for more options.

4) KITTI Completion

  1. Download KITTI data in the kitti folder on Google Drive.
  2. Run python3 test_kitti.py. Type python3 test_kitti.py -h for more options.

5) KITTI Registration

  1. Run the KITTI completion experiment first to get complete point clouds.
  2. Run python3 kitti_registration.py. Type python3 kitti_registration.py -h for more options.

6) Training

  1. Download training (train.lmdb, train.lmdb-lock) and validation (valid.lmdb, valid.lmdb-lock) data from shapenet or shapenet_car directory on Google Drive. Note that the training data for all 8 categories in shapenet takes up 49G of disk space. The training data for only the car category takes 9G instead.
  2. Run python3 train.py. Type python3 train.py -h for more options.

7) Data Generation

To generate your own data from ShapeNet, first Download ShapeNetCore.v1. Then, create partial point clouds from depth images (see instructions in render) and corresponding ground truths by sampling from CAD models (see instructions in sample). Finally, serialize the data using lmdb_writer.py.

License

This project Code is released under the MIT License (refer to the LICENSE file for details).

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].