DRC: Enhancing Personalized Image Generation via Disentangled Representation Composition
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916878701559808 |
|---|---|
| author | Xu, Yiyan Zheng, Wuqiang Wang, Wenjie Zhu, Fengbin Hu, Xinting Zhang, Yang Feng, Fuli Chua, Tat-Seng |
| author_facet | Xu, Yiyan Zheng, Wuqiang Wang, Wenjie Zhu, Fengbin Hu, Xinting Zhang, Yang Feng, Fuli Chua, Tat-Seng |
| contents | Personalized image generation has emerged as a promising direction in multimodal content creation. It aims to synthesize images tailored to individual style preferences (e.g., color schemes, character appearances, layout) and semantic intentions (e.g., emotion, action, scene contexts) by leveraging user-interacted history images and multimodal instructions. Despite notable progress, existing methods -- whether based on diffusion models, large language models, or Large Multimodal Models (LMMs) -- struggle to accurately capture and fuse user style preferences and semantic intentions. In particular, the state-of-the-art LMM-based method suffers from the entanglement of visual features, leading to Guidance Collapse, where the generated images fail to preserve user-preferred styles or reflect the specified semantics.
To address these limitations, we introduce DRC, a novel personalized image generation framework that enhances LMMs through Disentangled Representation Composition. DRC explicitly extracts user style preferences and semantic intentions from history images and the reference image, respectively, to form user-specific latent instructions that guide image generation within LMMs. Specifically, it involves two critical learning stages: 1) Disentanglement learning, which employs a dual-tower disentangler to explicitly separate style and semantic features, optimized via a reconstruction-driven paradigm with difficulty-aware importance sampling; and 2) Personalized modeling, which applies semantic-preserving augmentations to effectively adapt the disentangled representations for robust personalized generation. Extensive experiments on two benchmarks demonstrate that DRC shows competitive performance while effectively mitigating the guidance collapse issue, underscoring the importance of disentangled representation learning for controllable and effective personalized image generation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_17349 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DRC: Enhancing Personalized Image Generation via Disentangled Representation Composition Xu, Yiyan Zheng, Wuqiang Wang, Wenjie Zhu, Fengbin Hu, Xinting Zhang, Yang Feng, Fuli Chua, Tat-Seng Computer Vision and Pattern Recognition Information Retrieval Personalized image generation has emerged as a promising direction in multimodal content creation. It aims to synthesize images tailored to individual style preferences (e.g., color schemes, character appearances, layout) and semantic intentions (e.g., emotion, action, scene contexts) by leveraging user-interacted history images and multimodal instructions. Despite notable progress, existing methods -- whether based on diffusion models, large language models, or Large Multimodal Models (LMMs) -- struggle to accurately capture and fuse user style preferences and semantic intentions. In particular, the state-of-the-art LMM-based method suffers from the entanglement of visual features, leading to Guidance Collapse, where the generated images fail to preserve user-preferred styles or reflect the specified semantics. To address these limitations, we introduce DRC, a novel personalized image generation framework that enhances LMMs through Disentangled Representation Composition. DRC explicitly extracts user style preferences and semantic intentions from history images and the reference image, respectively, to form user-specific latent instructions that guide image generation within LMMs. Specifically, it involves two critical learning stages: 1) Disentanglement learning, which employs a dual-tower disentangler to explicitly separate style and semantic features, optimized via a reconstruction-driven paradigm with difficulty-aware importance sampling; and 2) Personalized modeling, which applies semantic-preserving augmentations to effectively adapt the disentangled representations for robust personalized generation. Extensive experiments on two benchmarks demonstrate that DRC shows competitive performance while effectively mitigating the guidance collapse issue, underscoring the importance of disentangled representation learning for controllable and effective personalized image generation. |
| title | DRC: Enhancing Personalized Image Generation via Disentangled Representation Composition |
| topic | Computer Vision and Pattern Recognition Information Retrieval |
| url | https://arxiv.org/abs/2504.17349 |