AnyStory: Towards Unified Single and Multiple Subject Personalization in Text-to-Image Generation
This addresses the challenge of personalized subject generation in text-to-image models, particularly for multiple subjects, but appears incremental as it builds on existing methods like ReferenceNet and CLIP.
The paper tackles the problem of generating personalized images with specific subjects in text-to-image generation, especially for multiple subjects, and proposes AnyStory, which achieves high-fidelity personalization for both single and multiple subjects without sacrificing subject fidelity.
Recently, large-scale generative models have demonstrated outstanding text-to-image generation capabilities. However, generating high-fidelity personalized images with specific subjects still presents challenges, especially in cases involving multiple subjects. In this paper, we propose AnyStory, a unified approach for personalized subject generation. AnyStory not only achieves high-fidelity personalization for single subjects, but also for multiple subjects, without sacrificing subject fidelity. Specifically, AnyStory models the subject personalization problem in an "encode-then-route" manner. In the encoding step, AnyStory utilizes a universal and powerful image encoder, i.e., ReferenceNet, in conjunction with CLIP vision encoder to achieve high-fidelity encoding of subject features. In the routing step, AnyStory utilizes a decoupled instance-aware subject router to accurately perceive and predict the potential location of the corresponding subject in the latent space, and guide the injection of subject conditions. Detailed experimental results demonstrate the excellent performance of our method in retaining subject details, aligning text descriptions, and personalizing for multiple subjects. The project page is at https://aigcdesigngroup.github.io/AnyStory/ .