Visual Question Answering on RailVQA-bench Static 1.0 (test)
75.74Overall ScoreRailVQA-CoM
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| RailVQA-CoMLMM=Qwen3-VL2026.03 | 75.74 | 86.4 | 74.3 | 75.3 | 79.3 | 72.8 | 87.9 | 67.8 | |
| LMM OnlyLMM=Qwen3-VL2026.03 | 74.81 | 85.2 | 76.3 | 74.9 | 78.2 | 71.2 | 81.1 | 64.9 | |
| RailVQA-CoMLMM=InternVL3.52026.03 | 74.08 | 82.4 | 76.5 | 75.8 | 77 | 74.4 | 82.1 | 65.2 | |
| Unadapted DetectorLMM=InternVL3.52026.03 | 73.2 | 81.55 | 74.4 | 71.7 | 76.6 | 68.9 | 79.5 | 62.4 | |
| Unadapted DetectorLMM=Qwen3-VL2026.03 | 72.2 | 86.1 | 70.7 | 73 | 75.5 | 68.1 | 81.7 | 60.1 | |
| RailVQA-CoMLMM=LLaVA-1.52026.03 | 71.48 | 74.9 | 70 | 70.5 | 72.5 | 69.7 | 78.8 | 63 | |
| Unadapted DetectorLMM=LLaVA-1.52026.03 | 71.1 | 75.81 | 69 | 70 | 71.7 | 69.1 | 79 | 61.8 | |
| LMM OnlyLMM=InternVL3.52026.03 | 71.04 | 78.1 | 74.5 | 73.6 | 75.1 | 73.3 | 71.7 | 61 | |
| LMM OnlyLMM=LLaVA-1.52026.03 | 69.82 | 72.2 | 68.1 | 69.1 | 70.9 | 67.7 | 76.1 | 60.8 | |
| RailVQA-CoMLMM=LLaMA3-Vision2026.03 | 67 | 80.74 | 63.8 | 66.7 | 70.5 | 63 | 74.2 | 58.6 | |
| Unadapted DetectorLMM=LLaMA3-Vision2026.03 | 66.57 | 80.56 | 65.5 | 66.3 | 68.5 | 63.5 | 73.4 | 58.2 | |
| LMM OnlyLMM=LLaMA3-Vision2026.03 | 65.72 | 81.68 | 67.2 | 65.4 | 67.2 | 62.1 | 70.5 | 57.7 |