AnyStory: Towards Unified Single and Multiple Subject Personalization in Text-to-Image Generation

TL;DR

AnyStory introduces a unified method for high-fidelity personalization in text-to-image generation, effectively handling both single and multiple subjects without compromising detail or accuracy.

Contribution

The paper presents a novel encode-then-route framework utilizing ReferenceNet and CLIP for personalized subject encoding and a decoupled router for precise subject placement in generated images.

Findings

Achieves high-fidelity personalization for single subjects.

Effectively handles multiple subjects without loss of detail.

Demonstrates superior alignment with text descriptions.

Abstract

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/ .