PyTorch PointCloud
non-profit
AI & ML interests
Deep Learning for 3D Point Cloud / LiDAR
Recent Activity
View all activity
Organization Card
PyTorch PointCloud (torch-pointcloud) is a collection of point cloud models, layers, utilities, datasets, transforms,
and reference training / benchmarking scripts that aim to reproduce SOTA models and help researchers and engineers reuse common layers.
Find all the torch-pointcloud models here.
models 134
torch-pointcloud/voxelnext.nuscenes.openpcdet
Object Detection • 8M • Updated
torch-pointcloud/utonia.pretrain.pointcept
0.1B • Updated
torch-pointcloud/utonia-lp.scannet20.pointcept
0.1B • Updated
torch-pointcloud/spvcnn-47gmacs.semantickitti.mit-han-lab
8.8M • Updated
torch-pointcloud/spvcnn-30gmacs.semantickitti.mit-han-lab
5.46M • Updated
torch-pointcloud/spvcnn-119gmacs.semantickitti.mit-han-lab
21.8M • Updated
torch-pointcloud/spunet-v1m1.scannet20.pointcept
39.2M • Updated
torch-pointcloud/sonata-lp.scannet20.fair
0.1B • Updated
torch-pointcloud/sonata-base.pretrain.fair
0.1B • Updated
torch-pointcloud/second.kitti.openpcdet
Object Detection • 5.33M • Updated
datasets 0
None public yet