RealGeneral: Unifying Visual Generation via Temporal In-Context Learning with Video Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lin, Yijing, Huang, Mengqi, Zhuang, Shuhan, Mao, Zhendong
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911065602785280
author Lin, Yijing
Huang, Mengqi
Zhuang, Shuhan
Mao, Zhendong
author_facet Lin, Yijing
Huang, Mengqi
Zhuang, Shuhan
Mao, Zhendong
contents Unifying diverse image generation tasks within a single framework remains a fundamental challenge in visual generation. While large language models (LLMs) achieve unification through task-agnostic data and generation, existing visual generation models fail to meet these principles. Current approaches either rely on per-task datasets and large-scale training or adapt pre-trained image models with task-specific modifications, limiting their generalizability. In this work, we explore video models as a foundation for unified image generation, leveraging their inherent ability to model temporal correlations. We introduce RealGeneral, a novel framework that reformulates image generation as a conditional frame prediction task, analogous to in-context learning in LLMs. To bridge the gap between video models and condition-image pairs, we propose (1) a Unified Conditional Embedding module for multi-modal alignment and (2) a Unified Stream DiT Block with decoupled adaptive LayerNorm and attention mask to mitigate cross-modal interference. RealGeneral demonstrates effectiveness in multiple important visual generation tasks, e.g., it achieves a 14.5% improvement in subject similarity for customized generation and a 10% enhancement in image quality for canny-to-image task. Project page: https://lyne1.github.io/realgeneral_web/; GitHub Link: https://github.com/Lyne1/RealGeneral
format Preprint
id arxiv_https___arxiv_org_abs_2503_10406
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle RealGeneral: Unifying Visual Generation via Temporal In-Context Learning with Video Models
Lin, Yijing
Huang, Mengqi
Zhuang, Shuhan
Mao, Zhendong
Computer Vision and Pattern Recognition
Artificial Intelligence
Unifying diverse image generation tasks within a single framework remains a fundamental challenge in visual generation. While large language models (LLMs) achieve unification through task-agnostic data and generation, existing visual generation models fail to meet these principles. Current approaches either rely on per-task datasets and large-scale training or adapt pre-trained image models with task-specific modifications, limiting their generalizability. In this work, we explore video models as a foundation for unified image generation, leveraging their inherent ability to model temporal correlations. We introduce RealGeneral, a novel framework that reformulates image generation as a conditional frame prediction task, analogous to in-context learning in LLMs. To bridge the gap between video models and condition-image pairs, we propose (1) a Unified Conditional Embedding module for multi-modal alignment and (2) a Unified Stream DiT Block with decoupled adaptive LayerNorm and attention mask to mitigate cross-modal interference. RealGeneral demonstrates effectiveness in multiple important visual generation tasks, e.g., it achieves a 14.5% improvement in subject similarity for customized generation and a 10% enhancement in image quality for canny-to-image task. Project page: https://lyne1.github.io/realgeneral_web/; GitHub Link: https://github.com/Lyne1/RealGeneral
title RealGeneral: Unifying Visual Generation via Temporal In-Context Learning with Video Models
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2503.10406