Paper ID: 2312.08195

Concept-centric Personalization with Large-scale Diffusion Priors

Pu Cao, Lu Yang, Feng Zhou, Tianrui Huang, Qing Song

Despite large-scale diffusion models being highly capable of generating diverse open-world content, they still struggle to match the photorealism and fidelity of concept-specific generators. In this work, we present the task of customizing large-scale diffusion priors for specific concepts as concept-centric personalization. Our goal is to generate high-quality concept-centric images while maintaining the versatile controllability inherent to open-world models, enabling applications in diverse tasks such as concept-centric stylization and image translation. To tackle these challenges, we identify catastrophic forgetting of guidance prediction from diffusion priors as the fundamental issue. Consequently, we develop a guidance-decoupled personalization framework specifically designed to address this task. We propose Generalized Classifier-free Guidance (GCFG) as the foundational theory for our framework. This approach extends Classifier-free Guidance (CFG) to accommodate an arbitrary number of guidances, sourced from a variety of conditions and models. Employing GCFG enables us to separate conditional guidance into two distinct components: concept guidance for fidelity and control guidance for controllability. This division makes it feasible to train a specialized model for concept guidance, while ensuring both control and unconditional guidance remain intact. We then present a null-text Concept-centric Diffusion Model as a concept-specific generator to learn concept guidance without the need for text annotations. Code will be available at https://github.com/PRIV-Creation/Concept-centric-Personalization.

Submitted: Dec 13, 2023