Visual Question Answering on SIXray and PIDray combined (test)
49.22Instance Location (IL)STING-BEE
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| STING-BEEevaluation format=multiple-choice2025.04 | 49.22 | 79.21 | 80.04 | 45.24 | 27.76 | 52.85 | 35.03 | 52.81 | |
| STING-BEEtraining=detailed captions2025.04 | 49.2 | 79.2 | 80 | 45.2 | 27.8 | 52.8 | 35 | 52.8 | |
| MiniGPTevaluation format=multiple-choice2025.04 | 43.22 | 60.21 | 54.79 | 28.16 | 19.64 | 32.85 | 29.74 | 36.62 | |
| LLaVa 1.5mode=Fine-tuned2025.04 | 35.2 | 59.4 | 77.8 | 36.1 | 15.9 | 48.2 | 25.3 | 45.1 | |
| Qwen2-VL2025.04 | 31.3 | 15.1 | 41.3 | 24.1 | 18.8 | 36.1 | 13.1 | 27.8 | |
| Florence-2evaluation format=multiple-choice2025.04 | 30.11 | 37.5 | 39.84 | 29.95 | 21.16 | 35.8 | 29.12 | 32.27 | |
| LLaVA 1.5evaluation format=multiple-choice2025.04 | 29.73 | 56.67 | 74.04 | 34.24 | 13.76 | 40.85 | 24.03 | 41.94 | |
| Llama 3.22025.04 | 26.5 | 29.3 | 36.3 | 20.8 | 11.6 | 25.4 | 13.3 | 23.5 | |
| STING-BEEtraining=limited captions2025.04 | 25.3 | 41.3 | 69.2 | 35.4 | 20.5 | 42.2 | 27.6 | 41.8 | |
| GPT-4o2025.04 | 21.5 | 16.2 | 36.3 | 29.6 | 25.2 | 19.02 | 18.3 | 31.3 |