Skip to content

Latest commit

 

History

History
42 lines (20 loc) · 900 Bytes

03. DRACO - Dense Reconstruction And Canonicalization of Objects.md

File metadata and controls

42 lines (20 loc) · 900 Bytes

DRACO - Dense Reconstruction And Canonicalization of Objects

A weakly supervised canonicalization pipeline that predicts object mask, depth, and canonical NOCS map.

Poster Video


Faculty Name

Rahul Sajnani
AadilMehdi Sanchawala
(have not listed all)

Research Area

Object level canonicalization, pose estimation, and reconstruction.

Type of Work

Algorithm.

Current State of work

Technology designed and implemented.

Potential Applications

Canonical reconstruction provides information to robots and allows them to manipulate and interact with objects in their surroundings.

Keywords

Deep Learning for Visual Perception, Semantic Scene Understanding, Learning Categories and Concepts.