Visual Hallucination Evaluation on POPE MS-COCO (Popular Sampling, val)
82.77AccuracyInstructBLIP-14B
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| InstructBLIP-14Bzero-shot=true, backbone=14B2023.09 | 82.77 | 76.27 | 95.13 | 84.66 | 62.37 | |
| DREAMLLM-7Bzero-shot=true, backbone=7B2023.09 | 80.07 | 75.74 | 88.47 | 81.61 | 58.4 | |
| MiniGPT-4-14Bzero-shot=true, backbone=14B2023.09 | 69.73 | 65.86 | 81.93 | 73.02 | 62.2 | |
| mPLUG-Owl-7Bzero-shot=true, backbone=7B2023.09 | 50.9 | 50.46 | 99.4 | 66.94 | 98.57 | |
| MMGPT-7Bzero-shot=true, backbone=7B2023.09 | 50 | 50 | 100 | 66.67 | 100 | |
| LLaVA-13Bzero-shot=true, backbone=13B2023.09 | 49.87 | 49.93 | 99.27 | 66.44 | 99.4 |