ajhamdi/vointcloud

Voint Cloud: Multi-View Point Cloud Representation for 3D Understanding (ICLR 2023)

33
/ 100
Emerging

This project offers a new way to analyze 3D models more accurately for tasks like classifying objects or segmenting their parts. It takes standard 3D point cloud data and processes it with multiple camera views, creating a richer 'Voint cloud' representation. Researchers and engineers working on 3D computer vision tasks can use this to improve the performance of their 3D understanding systems.

No commits in the last 6 months.

Use this if you are working with 3D point cloud data for tasks like object classification or part segmentation and want to achieve state-of-the-art accuracy by incorporating multi-view information.

Not ideal if you are looking for a standalone application for end-users without programming knowledge, as this is a research-oriented code library for developers.

3D-computer-vision point-cloud-analysis object-segmentation 3D-object-classification robotics-perception
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 6 / 25
Maturity 16 / 25
Community 11 / 25

How are scores calculated?

Stars

22

Forks

3

Language

License

MIT

Last pushed

May 02, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/ajhamdi/vointcloud"

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