Instructional Temporal Answer Grounding in Video on VehicleVQA (test)
85.31R@0.3CACR
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| CACRFamily=LVLM, Training Setting=supervised2026.06 | 85.31 | 76.7 | 69.33 | 72.11 | |
| MutualSLFamily=VLP, Training Setting=supervised2026.06 | 78.74 | 69.81 | 53.14 | 65.74 | |
| VTPSLFamily=VLP, Training Setting=supervised2026.06 | 74.15 | 67.15 | 54.59 | 64.51 | |
| Ask2LocFamily=LVLM, Training Setting=supervised2026.06 | 67.42 | 52.65 | 35.54 | 55.28 | |
| TimescopeFamily=LVLM, Training Setting=supervised2026.06 | 65.74 | 50.87 | 32.09 | 48.45 | |
| Time-R1Family=LVLM, Training Setting=supervised2026.06 | 59.06 | 44.27 | 29.36 | 44.28 | |
| TimeZeroFamily=LVLM, Training Setting=zero-shot2026.06 | 58.6 | 42.87 | 27.01 | 42.41 | |
| VTG-R1Family=LVLM, Training Setting=supervised2026.06 | 53.24 | 40.8 | 26.56 | 38.99 | |
| TFVTGFamily=LVLM, Training Setting=zero-shot2026.06 | 47.98 | 37.8 | 22.1 | 35.45 | |
| MoRSFamily=VLP, Training Setting=supervised2026.06 | 42.75 | 31.54 | 25.97 | 44.81 | |
| GPT-3.5 (CoT)Family=LLM-CoT, Training Setting=zero-shot2026.06 | 36.33 | 19 | 7 | 27.79 | |
| GPT-3.5Family=LLM, Training Setting=zero-shot2026.06 | 30.67 | 15 | 8.33 | 24.93 | |
| ACRMFamily=VLP, Training Setting=supervised2026.06 | 20.77 | 12.1 | 8.27 | 22.28 | |
| VSLBaseFamily=VLP, Training Setting=supervised2026.06 | 18.95 | 8.64 | 4.28 | 20.11 | |
| TMLGAFamily=VLP, Training Setting=supervised2026.06 | 17.69 | 8.79 | 3.46 | 16.54 | |
| VSLNetFamily=VLP, Training Setting=supervised2026.06 | 16.53 | 8.47 | 4.03 | 20.07 | |
| Qwen2.5-VL-7B-InstructFamily=LVLM, Training Setting=zero-shot2026.06 | 8.49 | 4.72 | 2.83 | 7.03 | |
| Random Mode2026.06 | 6.52 | 2.75 | 1.54 | 5.22 |