Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2504.19724 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915262999035904 |
|---|---|
| author | Wang, Haofan Xu, Yujia Li, Yimeng Li, Junchen Zhang, Chaowei Wang, Jing Yang, Kejia Chen, Zhibo |
| author_facet | Wang, Haofan Xu, Yujia Li, Yimeng Li, Junchen Zhang, Chaowei Wang, Jing Yang, Kejia Chen, Zhibo |
| contents | Although contemporary text-to-image generation models have achieved remarkable breakthroughs in producing visually appealing images, their capacity to generate precise and flexible typographic elements, especially non-Latin alphabets, remains constrained. To address these limitations, we start from an naive assumption that text understanding is only a sufficient condition for text rendering, but not a necessary condition. Based on this, we present RepText, which aims to empower pre-trained monolingual text-to-image generation models with the ability to accurately render, or more precisely, replicate, multilingual visual text in user-specified fonts, without the need to really understand them. Specifically, we adopt the setting from ControlNet and additionally integrate language agnostic glyph and position of rendered text to enable generating harmonized visual text, allowing users to customize text content, font and position on their needs. To improve accuracy, a text perceptual loss is employed along with the diffusion loss. Furthermore, to stabilize rendering process, at the inference phase, we directly initialize with noisy glyph latent instead of random initialization, and adopt region masks to restrict the feature injection to only the text region to avoid distortion of the background. We conducted extensive experiments to verify the effectiveness of our RepText relative to existing works, our approach outperforms existing open-source methods and achieves comparable results to native multi-language closed-source models. To be more fair, we also exhaustively discuss its limitations in the end. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_19724 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RepText: Rendering Visual Text via Replicating Wang, Haofan Xu, Yujia Li, Yimeng Li, Junchen Zhang, Chaowei Wang, Jing Yang, Kejia Chen, Zhibo Computer Vision and Pattern Recognition Although contemporary text-to-image generation models have achieved remarkable breakthroughs in producing visually appealing images, their capacity to generate precise and flexible typographic elements, especially non-Latin alphabets, remains constrained. To address these limitations, we start from an naive assumption that text understanding is only a sufficient condition for text rendering, but not a necessary condition. Based on this, we present RepText, which aims to empower pre-trained monolingual text-to-image generation models with the ability to accurately render, or more precisely, replicate, multilingual visual text in user-specified fonts, without the need to really understand them. Specifically, we adopt the setting from ControlNet and additionally integrate language agnostic glyph and position of rendered text to enable generating harmonized visual text, allowing users to customize text content, font and position on their needs. To improve accuracy, a text perceptual loss is employed along with the diffusion loss. Furthermore, to stabilize rendering process, at the inference phase, we directly initialize with noisy glyph latent instead of random initialization, and adopt region masks to restrict the feature injection to only the text region to avoid distortion of the background. We conducted extensive experiments to verify the effectiveness of our RepText relative to existing works, our approach outperforms existing open-source methods and achieves comparable results to native multi-language closed-source models. To be more fair, we also exhaustively discuss its limitations in the end. |
| title | RepText: Rendering Visual Text via Replicating |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2504.19724 |