Follow My Eyes: Backdoor Attacks on VLM-based Scanpath Prediction CCS: Security and privacy Mobile and wireless security CCS: Computing methodologies Machine learning
Diana Romero ∗,1 , Mutahar Ali ∗,1 , Momin Ahmad Khan ∗,2 , Habiba Farrukh 1 , Fatima Anwar 2 , and Salma Elmalaki 1 · 2026
Read Follow My Eyes: Backdoor Attacks on VLM-based Scanpath Prediction CCS: Security and privacy Mobile and wireless security CCS: Computing methodologies Machine learning by Diana Romero ∗,1 , Mutahar Ali ∗,1 , Momin Ahmad Khan ∗,2 , Habiba Farrukh 1 , Fatima Anwar 2 , and Salma Elmalaki 1 (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-2604-08766.txt.
Description
Diana Romero and five colleagues present the first backdoor attacks on scanpath models, where models forecast fixation sequences during visual search. A fixed path backdoor clusters tainted samples and proves easy to spot. They tune hostile training on each scene instead: spatial tricks send search to a foe chosen object, while time stretch lengthens search with extra gazes. Time stretch reaches up to 93.5 percent success from just 540 tainted samples, and no tested defense removes the backdoors without breaking output quality.
Catalog record
| CREATOR | Diana Romero ∗,1 , Mutahar Ali ∗,1 , Momin Ahmad Khan ∗,2 , Habiba Farrukh 1 , Fatima Anwar 2 , and Salma Elmalaki 1 |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 93170 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.08766