Lu Zhang

Adversarial Object Hallucination Attacks in Video-Language Models via Intermediate Feature Alignment

SAT-LDM: Provably Generalizable Image Watermarking for Latent Diffusion Models with Self-Augmented Training