Ego-view Accident Reason Answering
Ego-view accident reason answering is a multi-choice Video Question Answering (VQA) task. However, we target the question “What is the reason for the accident in this video?”. The performance is measured by the accuracy (Acc), i.e., the percentage of questions that are correctly answered.
Leaderboard
| Methods | Years | Acc (val.) | Acc (test.) | Params.(M) |
|---|---|---|---|---|
| HCRN | 2020 | 65.81 | 64.65 | 42 |
| ClipBERT | 2021 | 72.09 | 72.71 | 137 |
| VGT | 2022 | 68.40 | 68.66 | 143 |
| FrozenGQA | 2023 | 77.10 | 77.01 | 30 |
| CoVGT | 2023 | 81.70 | 79.97 | 159 |
| SeViLA | 2023 | 89.26 | 89.02 | 108 |
| X2VLM | 2023 | 76.40 | 75.79 | 362 |
| Mist | 2023 | 75.40 | 74.70 | 382 |
| BIMBA | 2025 | 48.90 | 48.40 | 156 |
| Qwen3VL | 2025 | 54.40 | 54.60 | - |
| VideoLLaMA2 | 2025 | - | 50.95 | - |
| VideoLLaMA2 w/pt | 2025 | - | 52.89 | - |
| VideoChat2 | 2025 | - | 49.56 | - |
| Video-LLaVA | 2025 | - | 43.63 | - |
| DriveMM | 2025 | - | 24.22 | - |
| iFinder | 2025 | - | 63.39 | - |
Submit Your Results
You can submit your metric values via the provided form. Furthermore, we would highly appreciate your contribution with clear links to relevant articles and code for more in-depth analysis.
Click here to submit your results:
Submit