Towards Flexible Multi-Modal Document Models

迈向灵活的多模式文档模型

Improving Table Structure Recognition with Visual-Alignment Sequential Coordinate Modeling

通过视觉对齐顺序坐标建模改进表格结构识别

Unifying Layout Generation with a Decoupled Diffusion Model

使用解耦扩散模型统一布局生成

Conditional Text Image Generation with Diffusion Models

使用扩散模型生成条件文本图像

Turning a CLIP Model into a Scene Text Detector

将 CLIP 模型转变为场景文本检测器

Unifying Vision, Text, and Layout for Universal Document Processing

统一视觉、文本和布局以实现通用文档处理

Modeling Entities as Semantic Points for Visual Information Extraction in the Wild

将实体建模为野外视觉信息提取的语义点

GeoLayoutLM: Geometric Pre-Training for Visual Information Extraction

GeoLayoutLM：视觉信息提取的几何预训练

Handwritten Text Generation from Visual Archetypes

从视觉原型生成手写文本

Towards Robust Tampered Text Detection in Document Image: New Dataset and New Solution

实现文档图像中稳健的篡改文本检测：新数据集和新解决方案

M6Doc: A Large-Scale Multi-Format, Multi-Type, Multi-Layout, Multi-Language, Multi-Annotation Category Dataset for Modern Document Layout Analysis

M6Doc：用于现代文档布局分析的大规模多格式、多类型、多布局、多语言、多注释类别数据集

Disentangling Writer and Character Styles for Handwriting Generation

解开书写者和字符风格以生成手写体