Lightweight Neural Framework for Robust 3D Volume and Surface Estimation from Multi-View Images
Diego E. Farchione, Ramzi Idoughi, Peter Wonka · 2026
Read Lightweight Neural Framework for Robust 3D Volume and Surface Estimation from Multi-View Images by Diego E. Farchione, Ramzi Idoughi, Peter Wonka (2026) 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-2606-23653.txt.
Description
Farchione, Idoughi, and Wonka estimate the volume and surface area of objects directly from a handful of photos, in a 2026 paper on computer vision. Such measurements serve coral reef monitoring, dietary analysis, and medical body measurement. Older pipelines need special scanners, manual work, or full mesh building, and they degrade badly when views are few or data are noisy.
The proposed framework is fully feed forward, meaning one pass through the network gives the answer. It predicts scale normalized volume and surface area, plus an uncertainty estimate for each, straight from multi view images. A graph based decoder fuses point cloud reconstructions with features aligned to each view, skipping slow iterative optimization. Experiments show it beats current methods, especially with few input images, across coral monitoring, dietary analysis, and anthropometry.
Catalog record
| CREATOR | Diego E. Farchione, Ramzi Idoughi, Peter Wonka |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 60 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.23653