Hallucination Detection on MHaluBench Image-to-Text Segment-level
90.44Hallucinatory Precision
Gemini-based Self-Check
Updated 1mo ago
Evaluation Results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
2025.06 | 90.44 | 71.08 | 79.6 | 57.35 | 83.8 | 68.1 | 75.11 | 73.89 | 77.44 | 73.85 | |
2025.06 | 89.53 | 93.47 | 91.46 | 84.38 | 76.33 | 80.15 | 88.06 | 86.95 | 84.9 | 85.8 | |
2025.06 | 89.3 | 47.71 | 62.19 | 43.76 | 87.68 | 58.38 | 60.38 | 66.53 | 67.69 | 60.29 | |
2025.06 | 88.77 | 78.76 | 83.46 | 63.17 | 78.52 | 70.02 | 78.68 | 75.97 | 78.64 | 76.74 | |
2025.06 | 87.03 | 91.01 | 88.98 | 78.52 | 70.77 | 74.44 | 84.6 | 82.77 | 80.89 | 81.71 | |
2025.06 | 82 | 79.98 | 80.98 | 76.04 | 78.35 | 77.18 | 79.25 | 79.02 | 79.16 | 79.08 | |
2025.06 | 79.37 | 74.17 | 76.68 | 70.52 | 76.22 | 73.26 | 75.09 | 74.94 | 75.19 | 74.97 |