arXiv 2602.122052026-02-13cs.CVcs.AI

DeepGen 1.0: 一种轻量级统一多模态模型，用于推进图像生成与编辑

DeepGen 1.0: A Lightweight Unified Multimodal Model for Advancing Image Generation and Editing

Shanghai Innovation Institute（上海创新研究院）；Fudan University（复旦大学）；University of Science and Technology of China（中国科学技术大学）；Shanghai Jiao Tong University（上海交通大学）；Zhejiang University（浙江大学）；Westlake University（西湖大学）；Nanjing University（南京大学）；University of Southern California（南加州大学）

Dianyi Wang, Ruihang Li, Feng Han, Chaofan Ma, Wei Song, Siyuan Wang, Yibin Wang, Yi Xin, Hongjian Liu, Zhixiong Zhang, Shengyuan Ding, Tianhang Wang, Zhenglin … 展开作者

Dianyi Wang, Ruihang Li, Feng Han, Chaofan Ma, Wei Song, Siyuan Wang, Yibin Wang, Yi Xin, Hongjian Liu, Zhixiong Zhang, Shengyuan Ding, Tianhang Wang, Zhenglin Cheng, Tao Lin, Cheng Jin, Kaicheng Yu, Jingjing Chen, Wenjie Wang, Zhongyu Wei, Jiaqi Wang

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

DeepGen 1.0通过轻量级统一多模态模型在图像生成与编辑领域实现高性能，采用SCB框架和数据驱动训练策略，超越大参数模型表现。

摘要：

当前用于图像生成和编辑的统一多模态模型通常依赖于庞大的参数规模（例如，>10B），导致训练成本和部署足迹过高。在本工作中，我们提出了DeepGen 1.0，这是一种轻量级的5B统一模型，其综合能力可与或超越许多更大的模型相媲美。为克服紧凑模型在语义理解和细粒度控制方面的限制，我们引入了堆叠通道桥接（SCB），一种深度对齐框架，该框架从多个VLM层中提取层次特征，并将其与可学习的“思考令牌”融合，为生成 backbone 提供结构化、具有推理能力的指导。我们进一步设计了一种以数据为中心的训练策略，涵盖三个渐进阶段：（1）在大规模图像-文本对和编辑三元组上进行对齐预训练，以同步VLM和DiT表示；（2）在高质量的生成、编辑和推理任务混合上进行联合监督微调，以培养全方位能力；（3）使用MR-GRPO强化学习，利用奖励函数和监督信号的混合，从而在生成质量和与人类偏好的契合度上获得显著提升，同时保持稳定的训练进度并避免视觉伪影。尽管仅在约5000万样本上进行训练，DeepGen 1.0在多样化的基准测试中均取得了领先表现，其在WISE基准上超越了800亿参数的HunyuanImage模型28%，在UniREditBench基准上超越了270亿参数的Qwen-Image-Edit模型37%。通过开源我们的训练代码、权重和数据集，我们提供了一种高效、高性能的替代方案，以民主化统一多模态研究。

英文摘要：

Current unified multimodal models for image generation and editing typically rely on massive parameter scales (e.g., >10B), entailing prohibitive training costs and deployment footprints. In this work, we present DeepGen 1.0, a lightweight 5B unified model that achieves comprehensive capabilities competitive with or surpassing much larger counterparts. To overcome the limitations of compact models in semantic understanding and fine-grained control, we introduce Stacked Channel Bridging (SCB), a deep alignment framework that extracts hierarchical features from multiple VLM layers and fuses them with learnable 'think tokens' to provide the generative backbone with structured, reasoning-rich guidance. We further design a data-centric training strategy spanning three progressive stages: (1) Alignment Pre-training on large-scale image-text pairs and editing triplets to synchronize VLM and DiT representations, (2) Joint Supervised Fine-tuning on a high-quality mixture of generation, editing, and reasoning tasks to foster omni-capabilities, and (3) Reinforcement Learning with MR-GRPO, which leverages a mixture of reward functions and supervision signals, resulting in substantial gains in generation quality and alignment with human preferences, while maintaining stable training progress and avoiding visual artifacts. Despite being trained on only ~50M samples, DeepGen 1.0 achieves leading performance across diverse benchmarks, surpassing the 80B HunyuanImage by 28% on WISE and the 27B Qwen-Image-Edit by 37% on UniREditBench. By open-sourcing our training code, weights, and datasets, we provide an efficient, high-performance alternative to democratize unified multimodal research.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。