hongxiaoy/ISO

[ECCV 2024] Monocular Occupancy Prediction for Scalable Indoor Scenes

27
/ 100
Experimental

This project helps professionals in robotics, virtual reality, or architectural visualization create detailed 3D models of indoor spaces from just a single 2D image. It takes an ordinary camera image of a room and generates a full 3D occupancy map, labeling objects like chairs, tables, and walls, even for parts of the scene not visible in the original photo. This tool is for researchers and developers working on AI models that need to understand and reconstruct indoor environments.

No commits in the last 6 months.

Use this if you need to generate a complete 3D reconstruction of an indoor scene, including hidden surfaces and object classifications, using only a single input image.

Not ideal if you require real-time 3D reconstruction for highly dynamic environments or if you don't have access to GPU hardware for processing.

3D-reconstruction indoor-mapping robotics-navigation virtual-reality-environments scene-understanding
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 8 / 25
Maturity 16 / 25
Community 3 / 25

How are scores calculated?

Stars

66

Forks

1

Language

Python

License

Apache-2.0

Last pushed

Sep 24, 2024

Commits (30d)

0

Get this data via API

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

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