AI.info

Shiguang Shan

Explore Shiguang Shan on AI.info.

• Contrastive Spectral Rectification: Test-Time Defense towards Zero-shot Adversarial Robustness of CLIP
• Semantic Mismatch and Perceptual Degradation: A New Perspective on Image Editing Immunity
• Towards Transferable Defense Against Malicious Image Edits
• JoPano: Unified Panorama Generation via Joint Modeling
• V-Attack: Targeting Disentangled Value Features for Controllable Adversarial Attacks on LVLMs
• Uniform Discrete Diffusion with Metric Path for Video Generation
• Revisiting Logit Distributions for Reliable Out-of-Distribution Detection