7 papers
DMAD: Distribution Matching as Adversarial Distillation for Fast Visual Generation
Zhengming Yu, Junkun Yuan, Haotian Yang +8
Distribution Matching Distillation (DMD) trains a few-step student from the difference between separately estimated target and student scores, so it must keep an auxiliary diffusio…
Token-Level Video Reinforcement Learning
Yifan Wang, Gordon Guocheng Qian, Yanyu Li +2
Reinforcement learning (RL) for video generation usually assigns one scalar reward to an entire sampled video. Yet a video is not uniformly flawed: some visual tokens may already s…
PDMD: Projected Distribution Matching Distillation for Video Diffusion Models
Zimo Wang, Junkun Yuan, Angtian Wang +9
Modern video diffusion models require tens of denoising evaluations over long spatiotemporal token sequences. Distribution Matching Distillation (DMD) reduces the number of functio…
EgoPlay: Event-Triggered Video Editing for Egocentric Streams
Jinjie Mai, Gordon Guocheng Qian, Willi Menapace +8
We introduce EgoPlay, an event-triggered video-to-video editor for egocentric streams, obtained by fine-tuning a pretrained V2V diffusion transformer on event-conditioned data buil…
Helix4D: Complex 4D Mesh Generation
Jiraphon Yenphraphai, Jianqi Chen, Jian Wang +6
Current video-to-4D methods struggle with complex topology changes, transparent materials, thin structures, and inner surfaces. We present Helix4D, a dynamic mesh generation framew…
FFAvatar: Few-Shot, Feed-Forward, and Generalizable Avatar Reconstruction
Thuan Hoang Nguyen, Jiahao Luo, Yinyu Nie +3
Avatar reconstruction has traditionally relied on per-subject optimization that requires hours of computation or on expensive preprocessing that limits scalability. We introduce FF…
Diffusion-DRF: Free, Rich, and Differentiable Reward for Video Diffusion Fine-Tuning
Yifan Wang, Yanyu Li, Gordon Guocheng Qian +3
Video diffusion alignment has been heavily relied on scalar rewards. These rewards are typically derived from learned reward models in human preference datasets, requiring addition…