SoDa 2 : Single-Stage Open-Set Domain Adaptation via Decoupled Alignment for Cross-Scene Hyperspectral Image Classification
Yiwen Liu, Minghua Wang, Jing Yao, Xin Zhao, Gemine Vivone · 2026
Read SoDa 2 : Single-Stage Open-Set Domain Adaptation via Decoupled Alignment for Cross-Scene Hyperspectral Image Classification by Yiwen Liu, Minghua Wang, Jing Yao, Xin Zhao, Gemine Vivone (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-2605-03371.txt.
Description
Yiwen Liu, Minghua Wang, Jing Yao, Xin Zhao, and Gemine Vivone present SoDa2, a single-stage method for classifying hyperspectral images across scenes when the target scene holds unknown classes. Mixed spectral-spatial features make direct alignment noisy, and two-stage training is costly.
The method splits spectral and spatial features, aligns each separately by discrepancy distance, and learns both aligned and free feature branches in one stage. A mixture model on their similarity separates known from unknown classes with no prior list needed. On three dataset groups it beat leading methods in accuracy and transfer for open-set cross-scene tasks.
Catalog record
| CREATOR | Yiwen Liu, Minghua Wang, Jing Yao, Xin Zhao, Gemine Vivone |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.03371