DREAM：基于显式自适应记忆的文档识别
DREAM: Document Recognition with Explicit Adaptive Memory
把显式的多尺度原型记忆作为即插即用模块接入编码器-解码器，让局部区域稀疏检索可解释的结构原型（如边框、倾斜文本），为文档/手写识别等视觉-语言任务提供显式区域级结构上下文，可迁移到任何需要布局先验的视觉编码器。
原文依据 A Memory Retrieval Mechanism enables local regions to sparsely attend to a few prototypes
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论大型多模态模型（LMM）在多种文档识别任务上已展现出良好性能。然而，LMM 采用隐式建模，其参数缺乏可解释性。受近期人类记忆与学习研究的启发，我们提出一种显式的多尺度原型记忆，用以增强文档识别模型，在不同空间分辨率上显式建模反复出现的版式与风格模式。记忆检索机制使局部区域能够稀疏地关注少量原型（例如图像边框、倾斜文本）；检索到的组合因子与视觉特征拼接后送入解码器，从而提供显式的区域级结构上下文。原型记忆巩固通过注意力加权的指数滑动平均（EMA）策略更新并稳定原型，同时稀疏性与抗坍缩正则化促进选择性激活。我们进一步采用分层记忆进行多分辨率编码。所提出的 DREAM 模块是一个即插即用组件，可无缝集成到多种编码器-解码器架构中。我们在两项任务上进行了验证，包括在公开数据集与自建的 DreamDoc 数据集上的文档识别，以及在 SCUT-HCCDoc 与 SCUT-EPT 数据集上的手写识别。实验结果表明所提方法是有效的。
英文原文摘要 1375 字符
Large multimodal models (LMMs) have shown promising performance for various document recognition tasks. However, LMMs adopt implicit modeling, and the parameters lack interpretability. Inspired by recent advances in human memory and learning research, we propose an explicit multiscale prototype memory that augments document recognition models, explicitly modeling recurrent layout and stylistic patterns across different spatial resolutions. A Memory Retrieval Mechanism enables local regions to sparsely attend to a few prototypes (e.g., image borders, tilted text); the retrieved compositional factors are concatenated with visual features and passed to the decoder, providing explicit region-wise structural context. Prototype memory consolidation updates and stabilizes prototypes via attention-weighted exponential moving average (EMA) strategy, while sparsity and anti-collapse regularization promote selective activation. We further adopt hierarchical memory for multi-resolution encoding. The proposed DREAM module is a plug-and-play component, allowing seamless integration into various encoder-decoder architectures. We validate on two tasks including document recognition on public datasets and the self-built DreamDoc dataset, and handwriting recognition on the SCUT-HCCDoc and SCUT-EPT datasets. Experimental results show that the proposed method is effective.