traveller59/spconv

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

2291 stars · 425 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

SpConv is a spatially sparse convolution library for deep learning on 3D point clouds and sparse tensors, distributed as PyPI packages with CPU and multiple CUDA builds. It provides fast sparse convolution primitives commonly used in LiDAR-based 3D object detection networks.

Use cases

• run sparse convolutions on 3D point clouds
• build lidar-based 3d object detection models
• process sparse voxel grids with deep learning
• accelerate 3d backbone networks for autonomous driving
• pip install sparse convolution with cuda support

When to choose

• you need high-performance spatial sparse convolution for point cloud networks
• you are building 3D detection or segmentation models on LiDAR data
• you want prebuilt wheels for specific CUDA versions on Linux

When to avoid

• you need dense 2D convolutions or general-purpose deep learning frameworks
• you require Windows or macOS support
• your project needs only CPU inference on non-Linux platforms

Facets

library · maturity active

deep-learning machine-learning image-processing deep-learning computer-vision autonomous-vehicles machine-learning python sparse-convolution point-cloud 3d-perception lidar pytorch tensorrt linux gpu cuda

Member repositories

For agents

markdown · JSON · MCP: product_card(name="traveller59/spconv")