FlowBlending: Stage-Aware Multi-Model Sampling for Fast and High-Fidelity Video Generation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Song, Jibin, Kwon, Mingi, Jeong, Jaeseok, Uh, Youngjung
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911347790315520
author Song, Jibin
Kwon, Mingi
Jeong, Jaeseok
Uh, Youngjung
author_facet Song, Jibin
Kwon, Mingi
Jeong, Jaeseok
Uh, Youngjung
contents In this work, we show that the impact of model capacity varies across timesteps: it is crucial for the early and late stages but largely negligible during the intermediate stage. Accordingly, we propose FlowBlending, a stage-aware multi-model sampling strategy that employs a large model and a small model at capacity-sensitive stages and intermediate stages, respectively. We further introduce simple criteria to choose stage boundaries and provide a velocity-divergence analysis as an effective proxy for identifying capacity-sensitive regions. Across LTX-Video (2B/13B) and WAN 2.1 (1.3B/14B), FlowBlending achieves up to 1.65x faster inference with 57.35% fewer FLOPs, while maintaining the visual fidelity, temporal coherence, and semantic alignment of the large models. FlowBlending is also compatible with existing sampling-acceleration techniques, enabling up to 2x additional speedup. Project page is available at: https://jibin86.github.io/flowblending_project_page.
format Preprint
id arxiv_https___arxiv_org_abs_2512_24724
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FlowBlending: Stage-Aware Multi-Model Sampling for Fast and High-Fidelity Video Generation
Song, Jibin
Kwon, Mingi
Jeong, Jaeseok
Uh, Youngjung
Computer Vision and Pattern Recognition
In this work, we show that the impact of model capacity varies across timesteps: it is crucial for the early and late stages but largely negligible during the intermediate stage. Accordingly, we propose FlowBlending, a stage-aware multi-model sampling strategy that employs a large model and a small model at capacity-sensitive stages and intermediate stages, respectively. We further introduce simple criteria to choose stage boundaries and provide a velocity-divergence analysis as an effective proxy for identifying capacity-sensitive regions. Across LTX-Video (2B/13B) and WAN 2.1 (1.3B/14B), FlowBlending achieves up to 1.65x faster inference with 57.35% fewer FLOPs, while maintaining the visual fidelity, temporal coherence, and semantic alignment of the large models. FlowBlending is also compatible with existing sampling-acceleration techniques, enabling up to 2x additional speedup. Project page is available at: https://jibin86.github.io/flowblending_project_page.
title FlowBlending: Stage-Aware Multi-Model Sampling for Fast and High-Fidelity Video Generation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.24724