DRoPS: D ynamic 3D R econstruction o f P re- S canned Objects
Narek Tumanyan, Samuel Rota Bulò, Denis Rozumny, Lorenzo Porzi, Adam Harley, Tali Dekel, Peter Kontschieder, Jonathon Luiten · 2026
Read DRoPS: D ynamic 3D R econstruction o f P re- S canned Objects by Narek Tumanyan, Samuel Rota Bulò, Denis Rozumny, Lorenzo Porzi, Adam Harley, Tali Dekel, Peter Kontschieder, Jonathon Luiten (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-2603-24770.txt.
Description
Tumanyan and co-authors present DRoPS, a method for dynamic 3D reconstruction of pre-scanned objects from casual video. Rebuilding moving scenes is ill-posed, and prior methods lean on distilled image priors plus hand-crafted motion rules. They struggle at extreme new viewpoints, above all with articulated motion. DRoPS uses a static pre-scan of the object as an explicit shape and appearance prior. Gaussian primitives sit in pixel grids anchored to the surface, and a CNN parametrises motion on those grids, so nearby points move together.
Catalog record
| CREATOR | Narek Tumanyan, Samuel Rota Bulò, Denis Rozumny, Lorenzo Porzi, Adam Harley, Tali Dekel, Peter Kontschieder, Jonathon Luiten |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 47678 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.24770