- Anomaly Detection with Foundation Models.- GPT-4V-AD: Exploring Grounding Potential of VQA-oriented GPT-4V for Zero-shot Anomaly Detection.- CLIP-AD: A Language-Guided Staged Dual-Path Model for Zero-shot Anomaly Detection.- DDPM-MoCo: Advancing Industrial Surface Defect Generation and Detection with Generative and Contrastive Learning.- Dual Memory-guided Probabilistic Model for Weakly-supervised Anomaly Detection.- Deep Learning for Human Activity Recognition.- Real-Time Human Action Prediction via Pose Kinematics.- Uncertainty Awareness for Unsupervised Domain Adaptation on Human Activity Recognition.

- Deep Interaction Feature Fusion for Robust Human Activity Recognition.- How effective are Self-Supervised models for Contact Identification in Videos.- A Wearable Multi-Modal Edge-Computing System for Real-Time Kitchen Activity Recognition.