FlashSLAM: Accelerated RGB-D SLAM for Real-Time 3D Scene Reconstruction with Gaussian Splatting flashslam.github.io
Phu Pham, Damon Conover, Aniket Bera · 2024
Read FlashSLAM: Accelerated RGB-D SLAM for Real-Time 3D Scene Reconstruction with Gaussian Splatting flashslam.github.io by Phu Pham, Damon Conover, Aniket Bera (2024) 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-2412-00682.txt.
Description
Phu Pham, Damon Conover, and Aniket Bera presented FlashSLAM in 2024, a fast system that maps a scene in three dimensions while tracking the camera. It builds on a method that places many soft blobs in the scene. Older versions were slow and lost track with sparse views or big camera moves.
FlashSLAM pairs that scene method with a fast vision tracker that matches features and aligns point clouds. Pose finding takes far less time than a known rival, with no costly repeated rendering. Tests in sparse and dense settings, both made-up and real, showed strong gains. Noise from depth sensors, as on phones, is also handled, which widens use.
Catalog record
| CREATOR | Phu Pham, Damon Conover, Aniket Bera |
|---|---|
| YEAR | 2024 |
| 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/2412-00682