arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2511.22982cs.CVcs.AI
收藏

Ovis-Image 技术报告

Ovis-Image Technical Report

Alibaba Group(阿里巴巴集团)

Guo-Hua Wang, Liangfu Cao, Tianyu Cui, Minghao Fu, Xiaohao Chen, Pengxin Zhan, Jianshan Zhao, Lan Li, Bowen Fu, Jiaqi Liu, Qing-Guo Chen

AI总结:

Ovis-Image 是一个70亿参数的文本到图像模型,通过优化的多模态主干和文本聚焦训练方法,在紧凑架构下实现与大型开源模型相当的文本渲染性能。

摘要:

我们介绍了Ovis-Image,一个70亿参数的文本到图像模型,专门优化以实现高质量的文本渲染,旨在在严格的计算约束下高效运行。基于我们之前的Ovis-U1框架,Ovis-Image集成了基于扩散的视觉解码器与更强的Ovis 2.5多模态主干,利用一种以文本为中心的训练管道,结合大规模预训练与精心设计的后训练细化。尽管其架构紧凑,Ovis-Image在文本渲染性能上与显著更大的开源模型如Qwen-Image相当,并接近闭源系统如Seedream和GPT4o。关键的是,该模型可以在单块高端GPU上部署,内存需求适中,缩小了前沿级文本渲染与实际部署之间的差距。我们的结果表明,结合强大的多模态主干与精心设计的文本聚焦训练配方,足以实现可靠的双语文本渲染,而无需依赖过大或专有模型。

英文摘要:

We introduce $\textbf{Ovis-Image}$, a 7B text-to-image model specifically optimized for high-quality text rendering, designed to operate efficiently under stringent computational constraints. Built upon our previous Ovis-U1 framework, Ovis-Image integrates a diffusion-based visual decoder with the stronger Ovis 2.5 multimodal backbone, leveraging a text-centric training pipeline that combines large-scale pre-training with carefully tailored post-training refinements. Despite its compact architecture, Ovis-Image achieves text rendering performance on par with significantly larger open models such as Qwen-Image and approaches closed-source systems like Seedream and GPT4o. Crucially, the model remains deployable on a single high-end GPU with moderate memory, narrowing the gap between frontier-level text rendering and practical deployment. Our results indicate that combining a strong multimodal backbone with a carefully designed, text-focused training recipe is sufficient to achieve reliable bilingual text rendering without resorting to oversized or proprietary models.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑