Autoregressive Video Generation beyond Next Frames Prediction

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Ren, Sucheng, Chen, Chen, Wang, Zhenbang, Song, Liangchen, Zhu, Xiangxin, Yuille, Alan, Yang, Yinfei, Lu, Jiasen
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866915520207388672
author Ren, Sucheng
Chen, Chen
Wang, Zhenbang
Song, Liangchen
Zhu, Xiangxin
Yuille, Alan
Yang, Yinfei
Lu, Jiasen
author_facet Ren, Sucheng
Chen, Chen
Wang, Zhenbang
Song, Liangchen
Zhu, Xiangxin
Yuille, Alan
Yang, Yinfei
Lu, Jiasen
contents Autoregressive models for video generation typically operate frame-by-frame, extending next-token prediction from language to video's temporal dimension. We question that unlike word as token is universally agreed in language if frame is a appropriate prediction unit? To address this, we present VideoAR, a unified framework that supports a spectrum of prediction units including full frames, key-detail frames, multiscale refinements, and spatiotemporal cubes. Among these designs, we find model video generation using \textit{spatiotemporal} cubes as prediction units, which allows autoregressive models to operate across both spatial and temporal dimensions simultaneously. This approach eliminates the assumption that frames are the natural atomic units for video autoregression. We evaluate VideoAR across diverse prediction strategies, finding that cube-based prediction consistently delivers superior quality, speed, and temporal coherence. By removing the frame-by-frame constraint, our video generator surpasses state-of-the-art baselines on VBench while achieving faster inference and enabling seamless scaling to minute-long sequences. We hope this work will motivate rethinking sequence decomposition in video and other spatiotemporal domains.
format Preprint
id arxiv_https___arxiv_org_abs_2509_24081
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Autoregressive Video Generation beyond Next Frames Prediction
Ren, Sucheng
Chen, Chen
Wang, Zhenbang
Song, Liangchen
Zhu, Xiangxin
Yuille, Alan
Yang, Yinfei
Lu, Jiasen
Computer Vision and Pattern Recognition
Autoregressive models for video generation typically operate frame-by-frame, extending next-token prediction from language to video's temporal dimension. We question that unlike word as token is universally agreed in language if frame is a appropriate prediction unit? To address this, we present VideoAR, a unified framework that supports a spectrum of prediction units including full frames, key-detail frames, multiscale refinements, and spatiotemporal cubes. Among these designs, we find model video generation using \textit{spatiotemporal} cubes as prediction units, which allows autoregressive models to operate across both spatial and temporal dimensions simultaneously. This approach eliminates the assumption that frames are the natural atomic units for video autoregression. We evaluate VideoAR across diverse prediction strategies, finding that cube-based prediction consistently delivers superior quality, speed, and temporal coherence. By removing the frame-by-frame constraint, our video generator surpasses state-of-the-art baselines on VBench while achieving faster inference and enabling seamless scaling to minute-long sequences. We hope this work will motivate rethinking sequence decomposition in video and other spatiotemporal domains.
title Autoregressive Video Generation beyond Next Frames Prediction
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2509.24081