Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/CVMI-Lab/Total-Decom
https://github.com/CVMI-Lab/Total-Decom
Last synced: about 1 month ago
JSON representation
- Host: GitHub
- URL: https://github.com/CVMI-Lab/Total-Decom
- Owner: CVMI-Lab
- Created: 2024-03-27T15:15:05.000Z (10 months ago)
- Default Branch: master
- Last Pushed: 2024-04-01T18:30:47.000Z (10 months ago)
- Last Synced: 2024-04-01T19:34:25.345Z (9 months ago)
- Size: 141 MB
- Stars: 21
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
- Awesome-Segment-Anything - [code
README
Total-Decom: Decomposed 3D Scene Reconstruction with Minimal Interaction
Xiaoyang Lyu*
·
Chirui Chang*
·
Peng Dai
·
Yang-Tian Sun
·
Xiaojuan Qi
*Equal Contributions
CVPR 2024
Paper | Project Page
TL; DR: Scene reconstruction from multi-view images is a fundamental problem in computer vision and graphics. Recent neural implicit surface reconstruction methods have achieved high-quality results; however, editing and manipulating the 3D geometry of reconstructed scenes remains challenging due to the absence of naturally decomposed object entities and complex object/background compositions. In this paper, we present Total-Decom, a novel method for decomposed 3D reconstruction with minimal human interaction. Our approach seamlessly integrates the Segment Anything Model (SAM) with hybrid implicit-explicit neural surface representations and a mesh-based region-growing technique for accurate 3D object decomposition. Total-Decom requires minimal human annotations while providing users with real-time control over the granularity and quality of decomposition. We extensively evaluate our method on benchmark datasets and demonstrate its potential for downstream applications, such as animation and scene editing.
# TODO
- [x] Create the project page
- [ ] Opensource all the training code
- [ ] Opensource the GUI
- [ ] Downstream applications