Event Localization and Captioning on ST-Align
60.4tIoU@0.5LLaVA-ST
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| LLaVA-STLLM Scale=7B2025.01 | 60.4 | 60 | 32.4 | 33.5 | 24.7 | |
| Grounded-VideoLLMLLM Scale=4B2025.01 | 53.1 | 56.4 | — | — | 7.2 | |
| VTimeLLMLLM Scale=7B2025.01 | 33.1 | 40.3 | — | — | 6 | |
| GroundingGPTLLM Scale=7B2025.01 | 4.8 | 2.1 | 6.4 | — | 8.2 |