Jiaju Han

Papers on Lattice

Total citations

Topics

Publication activitypapers/week, last 8 weeks

Research focus

Computer Vision (2)Multimodal Models (2)Red-Teaming & Adversarial Robustness (2)

Frequent co-authors

Xiang Chen (2)Chengyin Hu (2)Yiwei Wei (2)Jiahuan Long (2)

Papers (2)

May 21, 2026

Xiang Chen +83w ago

Exposing Vulnerabilities in Visible-Infrared VLMs: A Unified Geometric Adversarial Framework with Cross-Task Transferability

Even with robust training techniques like EOT, a carefully crafted adversarial patch can reliably fool VIS-IR VLMs and transfer across tasks like classification, captioning, and VQA.

Xiang Chen, Yuxian Dong, Chao Li +6

Computer Vision Multimodal Models Red-Teaming & Adversarial Robustness

Mar 30, 2026

Chengyin Hu +7Mar 30, 2026

XSPA: Crafting Imperceptible X-Shaped Sparse Adversarial Perturbations for Transferable Attacks on VLMs

VLMs can be devastatingly fooled by modifying less than 2% of image pixels in a fixed, X-shaped pattern, causing them to fail spectacularly across diverse tasks like classification, captioning, and question answering.

Chengyin Hu, Jiaju Han, Xuemeng Sun +5