Saved in:
Bibliographic Details
Main Authors: Li, Yaowei, Li, Xiaoyu, Zhang, Zhaoyang, Bian, Yuxuan, Liu, Gan, Li, Xinyuan, Xu, Jiale, Hu, Wenbo, Liu, Yating, Li, Lingen, Cai, Jing, Zou, Yuexian, He, Yancheng, Shan, Ying
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2507.01926
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918152053456896
author Li, Yaowei
Li, Xiaoyu
Zhang, Zhaoyang
Bian, Yuxuan
Liu, Gan
Li, Xinyuan
Xu, Jiale
Hu, Wenbo
Liu, Yating
Li, Lingen
Cai, Jing
Zou, Yuexian
He, Yancheng
Shan, Ying
author_facet Li, Yaowei
Li, Xiaoyu
Zhang, Zhaoyang
Bian, Yuxuan
Liu, Gan
Li, Xinyuan
Xu, Jiale
Hu, Wenbo
Liu, Yating
Li, Lingen
Cai, Jing
Zou, Yuexian
He, Yancheng
Shan, Ying
contents Image customization, a crucial technique for industrial media production, aims to generate content that is consistent with reference images. However, current approaches conventionally separate image customization into position-aware and position-free customization paradigms and lack a universal framework for diverse customization, limiting their applications across various scenarios. To overcome these limitations, we propose IC-Custom, a unified framework that seamlessly integrates position-aware and position-free image customization through in-context learning. IC-Custom concatenates reference images with target images to a polyptych, leveraging DiT's multi-modal attention mechanism for fine-grained token-level interactions. We propose the In-context Multi-Modal Attention (ICMA) mechanism, which employs learnable task-oriented register tokens and boundary-aware positional embeddings to enable the model to effectively handle diverse tasks and distinguish between inputs in polyptych configurations. To address the data gap, we curated a 12K identity-consistent dataset with 8K real-world and 4K high-quality synthetic samples, avoiding the overly glossy, oversaturated look typical of synthetic data. IC-Custom supports various industrial applications, including try-on, image insertion, and creative IP customization. Extensive evaluations on our proposed ProductBench and the publicly available DreamBench demonstrate that IC-Custom significantly outperforms community workflows, closed-source models, and state-of-the-art open-source approaches. IC-Custom achieves about 73\% higher human preference across identity consistency, harmony, and text alignment metrics, while training only 0.4\% of the original model parameters. Project page: https://liyaowei-stu.github.io/project/IC_Custom
format Preprint
id arxiv_https___arxiv_org_abs_2507_01926
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle IC-Custom: Diverse Image Customization via In-Context Learning
Li, Yaowei
Li, Xiaoyu
Zhang, Zhaoyang
Bian, Yuxuan
Liu, Gan
Li, Xinyuan
Xu, Jiale
Hu, Wenbo
Liu, Yating
Li, Lingen
Cai, Jing
Zou, Yuexian
He, Yancheng
Shan, Ying
Computer Vision and Pattern Recognition
Image customization, a crucial technique for industrial media production, aims to generate content that is consistent with reference images. However, current approaches conventionally separate image customization into position-aware and position-free customization paradigms and lack a universal framework for diverse customization, limiting their applications across various scenarios. To overcome these limitations, we propose IC-Custom, a unified framework that seamlessly integrates position-aware and position-free image customization through in-context learning. IC-Custom concatenates reference images with target images to a polyptych, leveraging DiT's multi-modal attention mechanism for fine-grained token-level interactions. We propose the In-context Multi-Modal Attention (ICMA) mechanism, which employs learnable task-oriented register tokens and boundary-aware positional embeddings to enable the model to effectively handle diverse tasks and distinguish between inputs in polyptych configurations. To address the data gap, we curated a 12K identity-consistent dataset with 8K real-world and 4K high-quality synthetic samples, avoiding the overly glossy, oversaturated look typical of synthetic data. IC-Custom supports various industrial applications, including try-on, image insertion, and creative IP customization. Extensive evaluations on our proposed ProductBench and the publicly available DreamBench demonstrate that IC-Custom significantly outperforms community workflows, closed-source models, and state-of-the-art open-source approaches. IC-Custom achieves about 73\% higher human preference across identity consistency, harmony, and text alignment metrics, while training only 0.4\% of the original model parameters. Project page: https://liyaowei-stu.github.io/project/IC_Custom
title IC-Custom: Diverse Image Customization via In-Context Learning
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2507.01926