---
title: "MagicTailor: Component-Controllable Personalization in Text-to-Image Diffusion Models"
authors:
  - "Donghao Zhou"
  - "Jiancheng Huang"
  - "Jinbin Bai"
  - "Jiaze Wang"
  - "Hao Chen"
  - "Guangyong Chen"
  - "Xiaowei Hu"
  - "Pheng-Ann Heng"
doi: "10.24963/ijcai.2025/1136"
url: "https://ijcai.org/proceedings/2025/1136"
pdf: "https://ijcai.org/proceedings/2025/1136.pdf"
venue: "Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence"
year: 2025
pages: "10225-10233"
track: "AI, Arts & Creativity"
keywords:
  - "Theory and philosophy of arts and creativity in AI systems: Autonomous creative or artistic AI"
  - "Methods and resources: Machine learning, deep learning, neural models, reinforcement learning"
  - "Application domains: Images, movies and visual arts"
  - "Theory and philosophy of arts and creativity in AI systems: Computational paradigms, architectures and models for creativity"
  - "Methods and resources: AI systems for ideation"
---

# MagicTailor: Component-Controllable Personalization in Text-to-Image Diffusion Models

Donghao Zhou, Jiancheng Huang, Jinbin Bai, Jiaze Wang, Hao Chen, Guangyong Chen, Xiaowei Hu, Pheng-Ann Heng

*Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence*, pages 10225-10233. DOI: [10.24963/ijcai.2025/1136](https://doi.org/10.24963/ijcai.2025/1136)

## Abstract

Text-to-image diffusion models can generate high-quality images but lack fine-grained control of visual concepts, limiting their creativity. Thus, we introduce component-controllable personalization, a new task that enables users to customize and reconfigure individual components within concepts. This task faces two challenges: semantic pollution, where undesired elements disrupt the target concept, and semantic imbalance, which causes disproportionate learning of the target concept and component. To address these, we design MagicTailor, a framework that uses Dynamic Masked Degradation to adaptively perturb unwanted visual semantics and Dual-Stream Balancing for more balanced learning of desired visual semantics. The experimental results show that MagicTailor achieves superior performance in this task and enables more personalized and creative image generation.