CSGO: Content-Style Composition in Text-to-Image Generation

ICLR 2025 Conference Submission283 Authors

Keywords: image generation, style transfer, stylized synthesis

TL;DR: We start by constructing a style transfer-specific dataset and then design a simple yet effective framework to address image-driven style transfer, text-driven stylized synthesis, and text editing-driven stylized synthesis tasks.

Abstract:

The diffusion model has shown exceptional capabilities in controlled image generation, which has further fueled interest in image style transfer. Existing works mainly focus on training free-based methods (e.g., image inversion) due to the scarcity of specific data. In this study, we present a data construction pipeline for content-style-stylized image triplets that generates and automatically cleanses stylized triplets. Based on this pipeline, we construct a dataset IMAGStyle, the first large-scale style transfer dataset containing 210k image triplets, available for the community to explore and research.Equipped with IMAGStyle, we propose a simple yet effective framework CSGO, a style transfer model based on end-to-end training, which explicitly decouples content and style features employing independent feature injection. Our CSGO implements image-driven style transfer, text-driven stylized synthesis, and text editing-driven stylized synthesis in the same model.
We conduct extensive experiments on CSGO to validate the effectiveness of synthetic stylized data for style control. Meanwhile, ablation experiments show the effectiveness of CSGO.

Primary Area: generative models

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 283