让视觉模型像人一样按语义重排图像元素，提升理解能力
DeepSeek-OCR 2: Visual Causal Flow
- 用因果推理动态重排图像视觉标记，突破固定扫描顺序
- 在复杂布局图像上实现更符合人类认知的顺序理解
- 适合研究视觉语言模型与认知启发式架构的学者
我们提出 DeepSeek-OCR 2，探索一种新型编码器 DeepEncoder V2 的可行性，该编码器能根据图像语义动态重排视觉标记。传统视觉语言模型（VLM）在将视觉标记输入大语言模型（LLM）时，始终采用固定的光栅扫描顺序（从左上到右下）和固定位置编码，这与人类视觉感知相悖。人类视觉在处理复杂布局图像时，会依据内在逻辑结构采取灵活且语义连贯的扫描方式，呈现因果驱动的序列化处理。受此认知机制启发，DeepEncoder V2 被设计为赋予编码器因果推理能力，使其在内容解释前智能重排视觉标记。本工作探索一种新范式：是否可通过两个级联的一维因果推理结构有效实现二维图像理解，从而提供一种具备真正二维推理潜力的新架构。代码与模型权重已公开于 http://github.com/deepseek-ai/DeepSeek-OCR-2。
原文摘要 · Abstract (English)
We present DeepSeek-OCR 2 to investigate the feasibility of a novel encoder-DeepEncoder V2-capable of dynamically reordering visual tokens upon image semantics. Conventional vision-language models (VLMs) invariably process visual tokens in a rigid raster-scan order (top-left to bottom-right) with fixed positional encoding when fed into LLMs. However, this contradicts human visual perception, which follows flexible yet semantically coherent scanning patterns driven by inherent logical structures. Particularly for images with complex layouts, human vision exhibits causally-informed sequential processing. Inspired by this cognitive mechanism, DeepEncoder V2 is designed to endow the encoder with causal reasoning capabilities, enabling it to intelligently reorder visual tokens prior to LLM-based content interpretation. This work explores a novel paradigm: whether 2D image understanding can be effectively achieved through two-cascaded 1D causal reasoning structures, thereby offering a new architectural approach with the potential to achieve genuine 2D reasoning. Codes and model weights are publicly accessible at http://github.com/deepseek-ai/DeepSeek-OCR-2.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。