GaussianSSC: Triplane-Guided Directional Gaussian Fields for 3D Semantic Completion
Ruiqi Xian, Jing Liang, He Yin, Xuewei Qi, Dinesh Manocha · 2026
Read GaussianSSC: Triplane-Guided Directional Gaussian Fields for 3D Semantic Completion by Ruiqi Xian, Jing Liang, He Yin, Xuewei Qi, Dinesh Manocha (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-2603-21487.txt.
Description
Xian and co-authors present GaussianSSC, a method for semantic scene completion from a single camera view. The task asks for a full 3D voxel map with labels, including hidden regions. GaussianSSC keeps the voxel grid but adds Gaussian-weighted image aggregation, which tightens how image features align to voxels. Each voxel then carries a learned Gaussian field that refines triplane features in a direction-aware way. On SemanticKITTI, the method improved both occupancy recall and precision plus semantic scores over prior best results.
Catalog record
| CREATOR | Ruiqi Xian, Jing Liang, He Yin, Xuewei Qi, Dinesh Manocha |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 37233 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.21487