Key Value Pair Extraction On Sibr
Metrics
key-value pair F1
Results
Performance results of various models on this benchmark
Model Name | key-value pair F1 | Paper Title | Repository |
|---|---|---|---|
| LiLT ([InfoXLM]_base) | 72.76 | LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding | - |
| PEneo (LiLT[InfoXLM]_base) | 82.36 | PEneo: Unifying Line Extraction, Line Grouping, and Entity Linking for End-to-end Document Pair Extraction | - |
| LayoutXLM | 70.45 | LayoutXLM: Multimodal Pre-training for Multilingual Visually-rich Document Understanding | - |
| PEneo (LayoutLMv3_base_chinese) | 82.52 | PEneo: Unifying Line Extraction, Line Grouping, and Entity Linking for End-to-end Document Pair Extraction | - |
| Donut | 17.26 | OCR-free Document Understanding Transformer | - |
| LayoutLMv3_base_chinese | 73.51 | LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking | - |
| PEneo (LayoutXLM_base) | 82.23 | PEneo: Unifying Line Extraction, Line Grouping, and Entity Linking for End-to-end Document Pair Extraction | - |
0 of 7 row(s) selected.