4 papers
BAgger: Backwards Aggregation for Mitigating Drift in Autoregressive Video Diffusion Models
Ryan Po, Eric Ryan Chan, Changan Chen +1
Autoregressive video models are promising for world modeling via next-frame prediction, but they suffer from exposure bias: a mismatch between training on clean contexts and infere…
Interspatial Attention for Efficient 4D Human Video Generation
Ruizhi Shao, Yinghao Xu, Yujun Shen +5
Generating photorealistic videos of digital humans in a controllable manner is crucial for a plethora of applications. Existing approaches either build on methods that employ templ…
SocialGen: Modeling Multi-Human Social Interaction with Language Models
Heng Yu, Juze Zhang, Changan Chen +4
Human interactions in everyday life are inherently social, involving engagements with diverse individuals across various contexts. Modeling these social interactions is fundamental…
The Language of Motion: Unifying Verbal and Non-verbal Language of 3D Human Motion
Changan Chen, Juze Zhang, Shrinidhi K. Lakshmikanth +5
Human communication is inherently multimodal, involving a combination of verbal and non-verbal cues such as speech, facial expressions, and body gestures. Modeling these behaviors…