Steering and Rectifying Latent Representation Manifolds in Frozen Multi-modal LLMs for Video Anomaly Detection

Documenti analoghi

• Generative Human-Object Interaction Detection via Differentiable Cognitive Steering of Multi-modal LLMs
  di: Cai, Zhaolin, et al.
  Pubblicazione: (2025)
• HeadHunt-VAD: Hunting Robust Anomaly-Sensitive Heads in MLLM for Tuning-Free Video Anomaly Detection
  di: Cai, Zhaolin, et al.
  Pubblicazione: (2025)
• MMHead: Towards Fine-grained Multi-modal 3D Facial Animation
  di: Wu, Sijing, et al.
  Pubblicazione: (2024)
• HiProbe-VAD: Video Anomaly Detection via Hidden States Probing in Tuning-Free Multimodal LLMs
  di: Cai, Zhaolin, et al.
  Pubblicazione: (2025)
• ManipShield: A Unified Framework for Image Manipulation Detection, Localization and Explanation
  di: Xu, Zitong, et al.
  Pubblicazione: (2025)