Mix-of-Show: Decentralized Low-Rank Adaptation for Multi-Concept Customization of Diffusion Models

TL;DR

Mix-of-Show introduces a decentralized framework for multi-concept customization of diffusion models, enabling high-fidelity composition of multiple concepts through novel embedding-decomposed LoRA and gradient fusion techniques.

Contribution

The paper presents ED-LoRA and gradient fusion methods for decentralized multi-concept customization, addressing conflicts and identity loss in diffusion models.

Findings

Supports theoretically limitless concept fusion.

Achieves high-fidelity composition of characters, objects, and scenes.

Effectively addresses attribute binding and missing object issues.

Abstract

Public large-scale text-to-image diffusion models, such as Stable Diffusion, have gained significant attention from the community. These models can be easily customized for new concepts using low-rank adaptations (LoRAs). However, the utilization of multiple concept LoRAs to jointly support multiple customized concepts presents a challenge. We refer to this scenario as decentralized multi-concept customization, which involves single-client concept tuning and center-node concept fusion. In this paper, we propose a new framework called Mix-of-Show that addresses the challenges of decentralized multi-concept customization, including concept conflicts resulting from existing single-client LoRA tuning and identity loss during model fusion. Mix-of-Show adopts an embedding-decomposed LoRA (ED-LoRA) for single-client tuning and gradient fusion for the center node to preserve the in-domain essence of single concepts and support theoretically limitless concept fusion. Additionally, we introduce regionally controllable sampling, which extends spatially controllable sampling (e.g., ControlNet and T2I-Adaptor) to address attribute binding and missing object problems in multi-concept sampling. Extensive experiments demonstrate that Mix-of-Show is capable of composing multiple customized concepts with high fidelity, including characters, objects, and scenes.