OmniConsistency: Learning Style-Agnostic Consistency from Paired Stylization Data

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Song, Yiren, Liu, Cheng, Shou, Mike Zheng
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909622001991680
author Song, Yiren
Liu, Cheng
Shou, Mike Zheng
author_facet Song, Yiren
Liu, Cheng
Shou, Mike Zheng
contents Diffusion models have advanced image stylization significantly, yet two core challenges persist: (1) maintaining consistent stylization in complex scenes, particularly identity, composition, and fine details, and (2) preventing style degradation in image-to-image pipelines with style LoRAs. GPT-4o's exceptional stylization consistency highlights the performance gap between open-source methods and proprietary models. To bridge this gap, we propose \textbf{OmniConsistency}, a universal consistency plugin leveraging large-scale Diffusion Transformers (DiTs). OmniConsistency contributes: (1) an in-context consistency learning framework trained on aligned image pairs for robust generalization; (2) a two-stage progressive learning strategy decoupling style learning from consistency preservation to mitigate style degradation; and (3) a fully plug-and-play design compatible with arbitrary style LoRAs under the Flux framework. Extensive experiments show that OmniConsistency significantly enhances visual coherence and aesthetic quality, achieving performance comparable to commercial state-of-the-art model GPT-4o.
format Preprint
id arxiv_https___arxiv_org_abs_2505_18445
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle OmniConsistency: Learning Style-Agnostic Consistency from Paired Stylization Data
Song, Yiren
Liu, Cheng
Shou, Mike Zheng
Computer Vision and Pattern Recognition
Diffusion models have advanced image stylization significantly, yet two core challenges persist: (1) maintaining consistent stylization in complex scenes, particularly identity, composition, and fine details, and (2) preventing style degradation in image-to-image pipelines with style LoRAs. GPT-4o's exceptional stylization consistency highlights the performance gap between open-source methods and proprietary models. To bridge this gap, we propose \textbf{OmniConsistency}, a universal consistency plugin leveraging large-scale Diffusion Transformers (DiTs). OmniConsistency contributes: (1) an in-context consistency learning framework trained on aligned image pairs for robust generalization; (2) a two-stage progressive learning strategy decoupling style learning from consistency preservation to mitigate style degradation; and (3) a fully plug-and-play design compatible with arbitrary style LoRAs under the Flux framework. Extensive experiments show that OmniConsistency significantly enhances visual coherence and aesthetic quality, achieving performance comparable to commercial state-of-the-art model GPT-4o.
title OmniConsistency: Learning Style-Agnostic Consistency from Paired Stylization Data
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2505.18445