Yunchao Wei

5 papers Beijing Jiaotong University

• A Unified Reasoning Framework for Holistic Zero-Shot Video Anomaly Analysis
• DCI: Dual-Conditional Inversion for Boosting Diffusion-Based Image Editing
• FlexVAR: Flexible Visual Autoregressive Modeling without Residual Prediction
• Martian World Model: Controllable Video Synthesis with Physically Accurate 3D Reconstructions
• PreFM: Online Audio-Visual Event Parsing via Predictive Future Modeling