TwoHead-SwinFPN: A Unified DL Architecture for Synthetic Manipulation, Detection and Localization in Identity Documents
Chan Naseeb, Adeel Ashraf Cheema, Hassan Sami, Tayyab Afzal, Muhammad Omair, Usman Habib · 2026
Read TwoHead-SwinFPN: A Unified DL Architecture for Synthetic Manipulation, Detection and Localization in Identity Documents by Chan Naseeb, Adeel Ashraf Cheema, Hassan Sami, Tayyab Afzal, Muhammad Omair, Usman Habib (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-2601-12895.txt.
Description
Chan Naseeb and five co authors build a deep learning system to guard identity documents. Generative AI models have escalated the threat of synthetic manipulation of ID documents. Face swapping and text inpainting attacks can fake a document with care. A fake that passes visual checks threatens banks, borders, and online verification.
Their architecture is called TwoHead-SwinFPN. It performs binary classification and precise localization of manipulated regions at the same time. Two heads share one backbone, so detection and localization inform each other. Knowing where a forgery sits helps human reviewers trust the verdict. Unified systems like this could harden document checks against AI era fraud.
Catalog record
| CREATOR | Chan Naseeb, Adeel Ashraf Cheema, Hassan Sami, Tayyab Afzal, Muhammad Omair, Usman Habib |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 35 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-12895