NVlabs/VoxFormer

Official PyTorch implementation of VoxFormer [CVPR 2023 Highlight]

44
/ 100
Emerging

This tool helps autonomous vehicle developers and robotics engineers create complete 3D maps of their surroundings using only standard 2D camera images. It takes raw camera feeds and an estimated depth map, then generates a detailed 3D scene where every visible and occluded object is identified and categorized. The result is a richer understanding of the environment, crucial for navigation and interaction.

1,178 stars. No commits in the last 6 months.

Use this if you need to understand the full 3D geometry and semantics of a scene, including occluded areas, from camera images alone, for applications like self-driving or robotic perception.

Not ideal if your application requires real-time processing with extremely low latency on resource-constrained hardware, or if you already have access to high-fidelity LiDAR or radar data for 3D sensing.

autonomous-driving robotics-perception 3d-scene-reconstruction computer-vision semantic-mapping
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 18 / 25

How are scores calculated?

Stars

1,178

Forks

95

Language

Python

License

Last pushed

Dec 07, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/NVlabs/VoxFormer"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.