Event-based Video Question Answering on EvQA-Sparse 200 Questions
66AccuracyFRT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| FRTFPS=24, Model Size=8B2025.12 | 66 | 18,352 | |
| FRTFPS=24, Model Size=32B2025.12 | 66 | 18,352 | |
| FRTFPS=8, Model Size=32B2025.12 | 65.5 | 7,233 | |
| FRTFPS=24, Model Size=2B2025.12 | 65 | 18,352 | |
| FRTFPS=4, Model Size=32B2025.12 | 64.5 | 3,643 | |
| FRTFPS=24, Model Size=4B2025.12 | 63.5 | 18,352 | |
| FRTFPS=2, Model Size=32B2025.12 | 61 | 1,889 | |
| FRTFPS=4, Model Size=8B2025.12 | 61 | 3,643 | |
| FRTFPS=8, Model Size=8B2025.12 | 60.5 | 7,233 | |
| FRTFPS=8, Model Size=4B2025.12 | 59.5 | 7,233 | |
| FRTFPS=2, Model Size=8B2025.12 | 57.5 | 1,889 | |
| FRTFPS=8, Model Size=2B2025.12 | 57 | 7,233 | |
| FRTFPS=1, Model Size=32B2025.12 | 56.5 | 1,103 | |
| FRTFPS=4, Model Size=4B2025.12 | 56.5 | 3,643 | |
| FRTFPS=2, Model Size=4B2025.12 | 56 | 1,889 | |
| FRTFPS=4, Model Size=2B2025.12 | 55.5 | 3,643 | |
| FRTFPS=0.1, Model Size=32B2025.12 | 52.5 | 614 | |
| FRTFPS=1, Model Size=8B2025.12 | 52 | 1,103 | |
| FRTFPS=2, Model Size=2B2025.12 | 50 | 1,889 | |
| ARTModel Size=32B2025.12 | 47.5 | 348 | |
| FRTFPS=1, Model Size=2B2025.12 | 46.5 | 1,103 | |
| FRTFPS=1, Model Size=4B2025.12 | 46 | 1,103 | |
| FRTFPS=0.1, Model Size=4B2025.12 | 45.5 | 614 | |
| FRTFPS=0.1, Model Size=8B2025.12 | 45.5 | 614 | |
| FRTFPS=0.1, Model Size=2B2025.12 | 42.5 | 614 | |
| ARTModel Size=2B2025.12 | 40.5 | 348 | |
| ARTModel Size=8B2025.12 | 36 | 348 | |
| ARTModel Size=4B2025.12 | 32.5 | 348 |