Key Information Extraction On Cord
評価指標
F1
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
モデル名 | F1 | Paper Title | Repository |
|---|---|---|---|
| LILT | 96.07 | LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding | |
| LayoutLMv2BASE | 94.95 | LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding | |
| LayoutMask (base) | 96.99 | LayoutMask: Enhance Text-Layout Interaction in Multi-modal Pre-training for Document Understanding | - |
| RORE (GeoLayoutLM) | 98.52 | Modeling Layout Reading Order as Ordering Relations for Visually-rich Document Understanding | |
| TPP (LayoutMask) | 96.92 | Reading Order Matters: Information Extraction from Visually-rich Documents by Token Path Prediction | |
| GeoLayoutLM | 97.97 | GeoLayoutLM: Geometric Pre-training for Visual Information Extraction | |
| LayoutLMv3 Large | 97.46 | LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking | |
| LayoutMask (large) | 97.19 | LayoutMask: Enhance Text-Layout Interaction in Multi-modal Pre-training for Document Understanding | - |
| LayoutLMv2LARGE | 96.01 | LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding |
0 of 9 row(s) selected.