Continual Few-shot Adaptation for Synthetic Fingerprint Detection
Joseph Geo Benjamin, Anil K. Jain, and Karthik Nandakumar · 2026
Read Continual Few-shot Adaptation for Synthetic Fingerprint Detection by Joseph Geo Benjamin, Anil K. Jain, and Karthik Nandakumar (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-2603-14632.txt.
Description
Benjamin, Jain and Nandakumar detect synthetic fingerprints under continual few-shot adaptation. Generative models now fake fingerprints well enough to threaten scanners. Injected fakes can slip into enrollment or login steps. A fixed detector overfits its training fakes and misses new generators. Their setup treats each new fake style as a small-shot update task. Cross-entropy and contrastive losses train the features. A few old-style samples replay during tuning to curb forgetting. Tests span several network backbones and generator types.
Catalog record
| CREATOR | Joseph Geo Benjamin, Anil K. Jain, and Karthik Nandakumar |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 32137 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.14632