Hierarchical Cell-Level Parallelism for Large Tables
Process table cells in parallel during document parsing so complex HTML tables are reconstructed accurately from spatial coordinates.
Source paper
Hao Feng, Shu Wei, Xiang Fei, Wei Shi, Yingdong Han, Lei Liao, Jinghui Lu, Binghong Wu, Qi Liu, Chunhui Lin, Jingqun Tang, Hao Liu, Can Huang
Generated by AI. The paper’s authors did not write these.
Process table cells in parallel during document parsing so complex HTML tables are reconstructed accurately from spatial coordinates.
Augment synthetic document rendering pipelines with vertical layouts so vision models learn to parse ancient and non-Manhattan scripts.
Introduce handwriting recognition capabilities to digital document parsers by fine-tuning models on annotated handwriting datasets.
This paper introduces Dolphin, a multimodal document image parsing model that employs an "analyze-then-parse" paradigm to extract structured content from documents. By first identifying layout elements (e.g., paragraphs, tables) and their reading order, and then using these elements as anchors for parallel content parsing with specialized prompts, Dolphin achieves state-of-the-art performance and high efficiency compared to existing end-to-end VLMs.
Dolphin uses a two-stage architecture sharing a unified Vision-Language Model (Swin Transformer encoder + mBART decoder). Stage 1 (Layout Analysis) detects element bounding boxes and types in reading order using a global image prompt. Stage 2 (Content Parsing) crops these elements and processes them in parallel using element-specific prompts (e.g., specific prompts for HTML tables vs. LaTeX formulas) to generate the final structured text.