Learning 3D-Gaussian Simulators from RGB Videos
Mikel Zhobro, A. René Geist, Georg Martius · 2025
Read Learning 3D-Gaussian Simulators from RGB Videos by Mikel Zhobro, A. René Geist, Georg Martius (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-2503-24009.txt.
Description
Mikel Zhobro, Rene Geist, and Georg Martius train in a 2025 paper a simulator that learns physics straight from multi view color video. Learned simulators promise real world dynamics without hand built models. Past ones lean on depth maps, particle tracks, or hand made features that cap their reach.
Their system, 3DGSim, fuses scene rebuilding, particle dynamics, and video rendering in one training loop. A splatting front end learns a latent particle view of each scene from the footage. A point transformer then forecasts how those particles move and interact. Joint training bakes physical traits into each particle's latent code. Rigid, elastic, and cloth behaviors all emerge, plus edits to unseen scenes.
Catalog record
| CREATOR | Mikel Zhobro, A. René Geist, Georg Martius |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 69 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-24009