Lihua Jing

7

Papers

42

Total Citations

Papers (7)

PAD: Patch-Agnostic Defense against Adversarial Patch Attacks

INTER: Mitigating Hallucination in Large Vision-Language Models by Interaction Guidance Sampling

EntropyMark: Towards More Harmless Backdoor Watermark via Entropy-based Constraint for Open-source Dataset Copyright Protection

Frequency Shuffling and Enhancement for Open Set Recognition

The Victim and The Beneficiary: Exploiting a Poisoned Model to Train a Clean Model on Poisoned Data

Attention-Imperceptible Backdoor Attacks on Vision Transformers

Revisiting Change Captioning from Self-supervised Global-Part Alignment