Coca-Splat: Collaborative Optimization for Camera Parameters and 3D Gaussians
Jiamin Wu, Hongyang Li, Xiaoke Jiang, Yuan YAO, Lei Zhang · 2025
Read Coca-Splat: Collaborative Optimization for Camera Parameters and 3D Gaussians by Jiamin Wu, Hongyang Li, Xiaoke Jiang, Yuan YAO, Lei Zhang (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-2504-00639.txt.
Description
Jiamin Wu and co-authors present Coca-Splat, a method for building three-dimensional scenes from just a few photos, in a 2025 paper. Most older methods need known camera poses before they can place scene parts in space. This method learns the camera settings and the scene parts at the same time, inside one network. Its design borrows from a known detection model that updates its guesses layer by layer.
The scene parts are soft blobs whose centers are cast onto each view by the learned camera settings. Those cast points link the two kinds of guess, so each can correct the other. Extra ray data ties the link more tightly. In tests on two large photo sets, the method beat both pose-based and pose-free rivals under the same pose-free rules.
Catalog record
| CREATOR | Jiamin Wu, Hongyang Li, Xiaoke Jiang, Yuan YAO, Lei Zhang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 58 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-00639