Salvato in:
Dettagli Bibliografici
Autori principali: Hu, Teng, Zhang, Jiangning, Su, Zihan, Yi, Ran
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2510.18775
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909862287376384
author Hu, Teng
Zhang, Jiangning
Su, Zihan
Yi, Ran
author_facet Hu, Teng
Zhang, Jiangning
Su, Zihan
Yi, Ran
contents Recent advances in video generation have made it possible to produce visually compelling videos, with wide-ranging applications in content creation, entertainment, and virtual reality. However, most existing diffusion transformer based video generation models are limited to low-resolution outputs (<=720P) due to the quadratic computational complexity of the attention mechanism with respect to the output width and height. This computational bottleneck makes native high-resolution video generation (1080P/2K/4K) impractical for both training and inference. To address this challenge, we present UltraGen, a novel video generation framework that enables i) efficient and ii) end-to-end native high-resolution video synthesis. Specifically, UltraGen features a hierarchical dual-branch attention architecture based on global-local attention decomposition, which decouples full attention into a local attention branch for high-fidelity regional content and a global attention branch for overall semantic consistency. We further propose a spatially compressed global modeling strategy to efficiently learn global dependencies, and a hierarchical cross-window local attention mechanism to reduce computational costs while enhancing information flow across different local windows. Extensive experiments demonstrate that UltraGen can effectively scale pre-trained low-resolution video models to 1080P and even 4K resolution for the first time, outperforming existing state-of-the-art methods and super-resolution based two-stage pipelines in both qualitative and quantitative evaluations.
format Preprint
id arxiv_https___arxiv_org_abs_2510_18775
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle UltraGen: High-Resolution Video Generation with Hierarchical Attention
Hu, Teng
Zhang, Jiangning
Su, Zihan
Yi, Ran
Computer Vision and Pattern Recognition
Recent advances in video generation have made it possible to produce visually compelling videos, with wide-ranging applications in content creation, entertainment, and virtual reality. However, most existing diffusion transformer based video generation models are limited to low-resolution outputs (<=720P) due to the quadratic computational complexity of the attention mechanism with respect to the output width and height. This computational bottleneck makes native high-resolution video generation (1080P/2K/4K) impractical for both training and inference. To address this challenge, we present UltraGen, a novel video generation framework that enables i) efficient and ii) end-to-end native high-resolution video synthesis. Specifically, UltraGen features a hierarchical dual-branch attention architecture based on global-local attention decomposition, which decouples full attention into a local attention branch for high-fidelity regional content and a global attention branch for overall semantic consistency. We further propose a spatially compressed global modeling strategy to efficiently learn global dependencies, and a hierarchical cross-window local attention mechanism to reduce computational costs while enhancing information flow across different local windows. Extensive experiments demonstrate that UltraGen can effectively scale pre-trained low-resolution video models to 1080P and even 4K resolution for the first time, outperforming existing state-of-the-art methods and super-resolution based two-stage pipelines in both qualitative and quantitative evaluations.
title UltraGen: High-Resolution Video Generation with Hierarchical Attention
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2510.18775