FAME: Feature Activation Map Explanation on Image Classification and Face Recognition
Xinyi Zhang, Manuel Günther · 2026
Read FAME: Feature Activation Map Explanation on Image Classification and Face Recognition by Xinyi Zhang, Manuel Günther (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-2605-12017.txt.
Description
Xinyi Zhang and Manuel Günther propose FAME, an explanation method combining activation mapping with perturbation ideas for image classification and face recognition. Class activation mapping assumes feature map locations reflect only underlying regions, an assumption that fails for deeper networks. FAME instead uses network gradients to manipulate the input image directly rather than applying fixed patches. Tests showed its attribution maps compete with state-of-the-art systems on both tasks.
Catalog record
| CREATOR | Xinyi Zhang, Manuel Günther |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 49851 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.12017