S²Flow: Towards Fast and Authentic Training-Free High-Resolution Video Generation
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (12) · 9693-9701
DOI: 10.1609/aaai.v40i12.37932 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Chaoqun Wang und 2 weiteren Autor:innen.
Abstract (Originalsprache)
Rectified flow models have shown strong potential in high-fidelity video generation, yet extending them to high-resolution remains challenging due to the high cost of full attention and error accumulation in the ODE-solving process. In this paper, we propose S^2Flow, a training-free framework that enables efficient and authentic high-resolution video generation by jointly exploring Flow-guided Sparse attention and Second-order ODE solution. Specifically, S^2Flow exploits and transfers the semantic and structural information from the low-resolution flow trajectory to guide the high-resolution flow in two aspects. First, S^2Flow dynamically captures the sparse patterns of the spatio-temporal attention maps from low-resolution videos to construct localized 3D windows, enabling efficient window attention in high-resolution inference. This can significantly reduce redundant computation while preserving contextual dependencies. Second, S^2Flow adopts a second-order ODE solver based on Taylor expansion, where the high-order derivative is approximated via central difference from the low-resolution flow, facilitating accurate high-resolution denoising. Extensive experiments on VBench dataset demonstrate that S^2Flow outperforms prior methods in both visual quality and inference speed, enabling 4x acceleration on 2560x1536 video generation.
Zitieren
APA
Wang, C., Min, S., & Yang, X. (2026). S²Flow: Towards Fast and Authentic Training-Free High-Resolution Video Generation. Proceedings of the AAAI Conference on Artificial Intelligence, 40(12), 9693-9701. https://doi.org/10.1609/aaai.v40i12.37932
Vancouver
Wang C, Min S, Yang X. S²Flow: Towards Fast and Authentic Training-Free High-Resolution Video Generation. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(12):9693-9701. doi:10.1609/aaai.v40i12.37932
BibTeX
@article{Wang2026,
author = {Wang, Chaoqun and Min, Shaobo and Yang, Xu},
title = {S²Flow: Towards Fast and Authentic Training-Free High-Resolution Video Generation},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {12},
pages = {9693--9701},
doi = {10.1609/aaai.v40i12.37932}
}
EUVIMED