Shaobin Zhuang

Video-GPT via Next Clip Diffusion

LinearSR: Unlocking Linear Attention for Stable and Efficient Image Super-Resolution

Convergence Theory of Decentralized Diffusion Models via Pseudo-Non-Markov Analysis

WeTok: Powerful Discrete Tokenization for High-Fidelity Visual Reconstruction

UniFlow: A Unified Pixel Flow Tokenizer for Visual Understanding and Generation

Efficiently Access Diffusion Fisher: Within the Outer Product Span Space

TimeStep Master: Asymmetrical Mixture of Timestep LoRA Experts for Versatile and Efficient Diffusion Models in Vision

TimeStep Master: Asymmetrical Mixture of Timestep LoRA Experts for Versatile and Efficient Diffusion Models in Vision

Gradient-Free Analytical Fisher Information of Diffused Distributions