Chengwei Wang

Undergrad student, SUN YAT-SEN UNIVERSITY

1 paper at NeurIPS 2025

Homepage· OpenReview· Semantic Scholar· Google Scholar

Poster Session 5

1 paper

Friday, December 5, 2025 · 11:00 AM → 2:00 PM

Exhibit Hall C,D,E

Towards Irreversible Attack: Fooling Scene Text Recognition via Multi-Population Coevolution Search

#3909 · Jingyu Li, Pengwen Dai, Mingqing Zhu, Chengwei Wang, Haolong Liu, Xiaochun Cao

We propose a pixel-level black-box attack method to fool STR models to predict more incorrect characters, using a novel multi-population coevolution search algorithm.