Visual Information Extraction on POIE
94.45AccuracyQwen2-VL-7B + RIDGE
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Qwen2-VL-7B + RIDGEDataset Combinations=+RIDGE2025.04 | 94.45 | 96.71 | |
| Qwen2-VL-7B + RIDGE + DocVQA + FUNSDDataset Combinations=+RIDGE+DocVQA+FUNSD2025.04 | 94.11 | 96.72 | |
| Qwen2-VL-7B + RIDGE + DocVQADataset Combinations=+RIDGE+DocVQA2025.04 | 94.08 | 96.67 | |
| Qwen2-VL-7B+zero-shot=true, RIDGE fine-tuned=true2025.04 | 94.04 | 96.71 | |
| Qwen2-VL-7BDataset Combinations=Base Model2025.04 | 93.51 | 96.01 | |
| Qwen2-VL-7Bzero-shot=true, RIDGE fine-tuned=false2025.04 | 93.51 | 96.01 | |
| DocOwl-1.5-Chatzero-shot=true, RIDGE fine-tuned=false2025.04 | 51.57 | 61.35 | |
| LLaVA-NeXT-Mistral-7B+zero-shot=true, RIDGE fine-tuned=true2025.04 | 51.46 | 62.13 | |
| LLaVA-NeXT-Mistral-7Bzero-shot=true, RIDGE fine-tuned=false2025.04 | 49.88 | 60.05 | |
| Monkeyzero-shot=true, RIDGE fine-tuned=false2025.04 | 30.64 | 43.95 |