[2603.21864] Adaptive Video Distillation: Mitigating Oversaturation and Temporal Collapse in Few-Step Generation

Computer Science > Computer Vision and Pattern Recognition arXiv:2603.21864 (cs) [Submitted on 23 Mar 2026] Title:Adaptive Video Distillation: Mitigating Oversaturation and Temporal Collapse in Few-Step Generation Authors:Yuyang You, Yongzhi Li, Jiahui Li, Yadong Mu, Quan Chen, Peng Jiang View a PDF of the paper titled Adaptive Video Distillation: Mitigating Oversaturation and Temporal Collapse in Few-Step Generation, by Yuyang You and 5 other authors View PDF HTML (experimental) Abstract:Video generation has recently emerged as a central task in the field of generative AI. However, the substantial computational cost inherent in video synthesis makes model distillation a critical technique for efficient deployment. Despite its significance, there is a scarcity of methods specifically designed for video diffusion models. Prevailing approaches often directly adapt image distillation techniques, which frequently lead to artifacts such as oversaturation, temporal inconsistency, and mode collapse. To address these challenges, we propose a novel distillation framework tailored specifically for video diffusion models. Its core innovations include: (1) an adaptive regression loss that dynamically adjusts spatial supervision weights to prevent artifacts arising from excessive distribution shifts; (2) a temporal regularization loss to counteract temporal collapse, promoting smooth and physically plausible sampling trajectories; and (3) an inference-time frame interpolation strategy ...