8 papers
Learning How the World Evolves: Extrapolative Video World Models via Latent Dynamics Reasoning
Haodong Li, Shaoteng Liu, Tianyu Wang +7
The world evolves following its dynamics, i.e., its laws of motion. However, leading video diffusion models largely fit the pixels without modeling how the pixels transit over time…
SURF: Signature-Retained Fast Video Generation
Kaixin Ding, Xi Chen, Sihui Ji +4
The demand for high-resolution video generation is growing rapidly. However, the generation resolution is severely constrained by slow inference speeds. For instance, Wan2.1 requir…
MemFlow: Flowing Adaptive Memory for Consistent and Efficient Long Video Narratives
Sihui Ji, Xi Chen, Shuai Yang +3
The core challenge for streaming video generation is maintaining the content consistency in long context, which poses high requirement for the memory design. Most existing solution…
PhysMaster: Mastering Physical Representation for Video Generation via Reinforcement Learning
Sihui Ji, Xi Chen, Xin Tao +2
Video generation models nowadays are capable of generating visually realistic videos, but often fail to adhere to physical laws, limiting their ability to generate physically plaus…
LayerFlow: A Unified Model for Layer-aware Video Generation
Sihui Ji, Hao Luo, Xi Chen +3
We present LayerFlow, a unified solution for layer-aware video generation. Given per-layer prompts, LayerFlow generates videos for the transparent foreground, clean background, and…
DiffDoctor: Diagnosing Image Diffusion Models Before Treating
Yiyang Wang, Xi Chen, Xiaogang Xu +4
In spite of recent progress, image diffusion models still produce artifacts. A common solution is to leverage the feedback provided by quality assessment systems or human annotator…
VideoAnydoor: High-fidelity Video Object Insertion with Precise Motion Control
Yuanpeng Tu, Hao Luo, Xi Chen +3
Despite significant advancements in video generation, inserting a given object into videos remains a challenging task. The difficulty lies in preserving the appearance details of t…
FashionComposer: Compositional Fashion Image Generation
Sihui Ji, Yiyang Wang, Xi Chen +3
We present FashionComposer for compositional fashion image generation. Unlike previous methods, FashionComposer is highly flexible. It takes multi-modal input (i.e., text prompt, p…