Haiyang-W/DSVT
A sparse voxel transformer backbone for efficient 3D object detection from LiDAR point clouds.
Not currently ranked — collecting fresh signals.
star history
DSVT implements a dynamic sparse voxel transformer with rotated sets for processing large-scale point clouds. Published at CVPR 2023, it serves as an efficient and deployment-friendly sparse backbone for 3D object detection tasks. The model achieves state-of-the-art results on Waymo and nuScenes datasets for vehicle, pedestrian, and cyclist detection.
Frequently asked
- What is Haiyang-W/DSVT?
- A sparse voxel transformer backbone for efficient 3D object detection from LiDAR point clouds.
- Is DSVT open source?
- Yes — Haiyang-W/DSVT is open source, released under the Apache-2.0 license.
- What language is DSVT written in?
- Haiyang-W/DSVT is primarily written in Python.
- How popular is DSVT?
- Haiyang-W/DSVT has 454 stars on GitHub.
- Where can I find DSVT?
- Haiyang-W/DSVT is on GitHub at https://github.com/Haiyang-W/DSVT.