All Projects → hyangwinter → Flownet3d_pytorch

hyangwinter / Flownet3d_pytorch

The pytorch implementation of flownet3d: https://github.com/xingyul/flownet3d

Programming Languages

python
139335 projects - #7 most used programming language

Projects that are alternatives of or similar to Flownet3d pytorch

Awesome Robotics
A curated list of awesome links and software libraries that are useful for robots.
Stars: ✭ 478 (+1125.64%)
Mutual labels:  point-cloud
Pcx
Point cloud importer & renderer for Unity
Stars: ✭ 806 (+1966.67%)
Mutual labels:  point-cloud
Hdl graph slam
3D LIDAR-based Graph SLAM
Stars: ✭ 945 (+2323.08%)
Mutual labels:  point-cloud
3dmatch Toolbox
3DMatch - a 3D ConvNet-based local geometric descriptor for aligning 3D meshes and point clouds.
Stars: ✭ 571 (+1364.1%)
Mutual labels:  point-cloud
Lidar camera calibration
ROS package to find a rigid-body transformation between a LiDAR and a camera for "LiDAR-Camera Calibration using 3D-3D Point correspondences"
Stars: ✭ 734 (+1782.05%)
Mutual labels:  point-cloud
Gpdnet
Learning Graph-Convolutional Representations for Point Cloud Denoising (ECCV 2020)
Stars: ✭ 19 (-51.28%)
Mutual labels:  point-cloud
Draco
Draco is a library for compressing and decompressing 3D geometric meshes and point clouds. It is intended to improve the storage and transmission of 3D graphics.
Stars: ✭ 4,611 (+11723.08%)
Mutual labels:  point-cloud
Blender kitti
Render large point clouds and voxel grids with blender.
Stars: ✭ 36 (-7.69%)
Mutual labels:  point-cloud
3d Machine Learning
A resource repository for 3D machine learning
Stars: ✭ 7,405 (+18887.18%)
Mutual labels:  point-cloud
House Of Cards
Display of Radiohead's House of Cards point cloud data
Stars: ✭ 12 (-69.23%)
Mutual labels:  point-cloud
Cilantro
A lean C++ library for working with point cloud data
Stars: ✭ 577 (+1379.49%)
Mutual labels:  point-cloud
Pcl
Point Cloud Library (PCL)
Stars: ✭ 6,897 (+17584.62%)
Mutual labels:  point-cloud
3d Bounding Boxes From Monocular Images
A two stage multi-modal loss model along with rigid body transformations to regress 3D bounding boxes
Stars: ✭ 24 (-38.46%)
Mutual labels:  point-cloud
Superpoint graph
Large-scale Point Cloud Semantic Segmentation with Superpoint Graphs
Stars: ✭ 533 (+1266.67%)
Mutual labels:  point-cloud
Mmdetection3d
OpenMMLab's next-generation platform for general 3D object detection.
Stars: ✭ 945 (+2323.08%)
Mutual labels:  point-cloud
Bpy
blender python scripts
Stars: ✭ 441 (+1030.77%)
Mutual labels:  point-cloud
Pointcloudutilities
Utilities for point cloud processing. read ply, write ply, search nearest neighbors using octree ...
Stars: ✭ 17 (-56.41%)
Mutual labels:  point-cloud
Unity 3mx
Load 3MX/3MXB format LOD model files generated by Bentley ContextCapture into Unity.
Stars: ✭ 38 (-2.56%)
Mutual labels:  point-cloud
Point Cloud Filter
Scripts showcasing filtering techniques applied to point cloud data.
Stars: ✭ 34 (-12.82%)
Mutual labels:  point-cloud
Pointnet2 pytorch
PyTorch implementation of Pointnet2/Pointnet++
Stars: ✭ 844 (+2064.1%)
Mutual labels:  point-cloud

flownet3d_pytorch

The pytorch implementation of flownet3d based on WangYueFt/dcp, sshaoshuai/Pointnet2.PyTorch and yanx27/Pointnet_Pointnet2_pytorch

Installation

Requirements

PyTorch>=1.0: https://pytorch.org

scipy>=1.2

numpy

h5py

tqdm

Install

Install this library by running the following command:

cd lib
python setup.py install
cd ../

Training

The processed Flyingthings3d data is provided here for download (total size ~11GB).

Then run the following command to train:

python main.py --exp_name=flownet3d --dataset_path=xx/yy

where xx/yy is the dataset path

Performance comparison

All of the following experiments were tested on a TITAN RTX

  1. GPU memory usage:
batch size flownet3d(GB) flownet3d_pytorch(GB)
16 16.9 9.1
  1. Training time per epoch on Flyingthings3d dataset:
batch size flownet3d(min) flownet3d_pytorch(min)
16 6.7 3.4
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].