We are an AI Research Team at Zhejiang University

Projects

• DICEPTION: A Generalist Diffusion Model for Vision Perception

• Framer: Interactive frame interpolation

• 3D reconstruction using pre-trained depth models

• Generating comics using generative models

• Generating long videos using generative models

• GenDeF: Learning generative deformation field for video generation

• ACTIVE-O3: Empowering Multimodal Large Language Models with Active Perception via GRPO

• FreerCustom: Training-Free Multi-Concept Customization for Image and Video Generation

• Omni-R1: Reinforcement Learning for Omnimodal Reasoning via Two-System Collaboration

• SegAgent: Exploring Pixel Understanding Capabilities in MLLMs by Imitating Human Annotator Trajectories

• Revisiting Depth Representations for Feed-Forward 3D Gaussian Splatting