Statistical Confidence Rescoring for Robust 3D Scene Graph Generation from Multi-View Images
Qi Xun Yeo, Yanyan Li, Gim Hee Lee · 2025
Read Statistical Confidence Rescoring for Robust 3D Scene Graph Generation from Multi-View Images by Qi Xun Yeo, Yanyan Li, Gim Hee Lee (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-06546.txt.
Description
Qi Xun Yeo, Yanyan Li, and Gim Hee Lee build 3D scene graphs from photos alone in a 2025 computer vision paper. A scene graph lists objects in a space and how they relate. Most methods lean on true 3D labels, which are costly to make.
Their method starts from several views of a room and rebuilds rough 3D shape from predicted depth. Semantic masks strip out background noise from the features. Neighbor data enrich each object and link. A final rescoring step uses counts from training data to fix doubtful calls. Tests show the method beats other photo-only systems, closing part of the gap to label-fed ones.
Catalog record
| CREATOR | Qi Xun Yeo, Yanyan Li, Gim Hee Lee |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 47 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-06546