White-Box m HC: Electromagnetic Spectrum–Aware and Interpretable Stream Interactions for Hyperspectral Image Classification
Yimin Zhu, Lincoln Linlin Xu, Zhengsen Xu, Zack Dewis, Mabel Heffring, Saeid Taleghanidoozdoozan, Motasem Alkayid, Quinn Ledingham, Megan Greenwood · 2026
Read White-Box m HC: Electromagnetic Spectrum–Aware and Interpretable Stream Interactions for Hyperspectral Image Classification by Yimin Zhu, Lincoln Linlin Xu, Zhengsen Xu, Zack Dewis, Mabel Heffring, Saeid Taleghanidoozdoozan, Motasem Alkayid, Quinn Ledingham, Megan Greenwood (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-2601-15757.txt.
Description
Yimin Zhu and a large team make hyperspectral image classification easier to read in a 2026 arXiv paper. Hyperspectral images record many wavelength bands per pixel. Deep models classify them well, but their feature mixing is opaque. Users cannot see how a decision was reached. The team presents a white-box framework called ES-mHC. It is aware of the electromagnetic spectrum by design.
The framework models interactions among spectrum groupings directly. It uses structured, directional matrices for those interactions. The residual stream carries information between groupings in a traceable way. Interpretability is the headline gain. Analysts in remote sensing need to defend their maps. A model that shows its spectral reasoning earns more trust than a black box.
Catalog record
| CREATOR | Yimin Zhu, Lincoln Linlin Xu, Zhengsen Xu, Zack Dewis, Mabel Heffring, Saeid Taleghanidoozdoozan, Motasem Alkayid, Quinn Ledingham, Megan Greenwood |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 27 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-15757