Open-Vocabulary Audio-Visual Event Localization on OV-AVEBench (unseen)
65.6AccuracyHSCHG
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| HSCHG2026.06 | 65.6 | 55.7 | 48.2 | 56.5 | |
| fine-tuning baselinestrategy=fine-tuning2024.11 | 64.9 | 55 | 47.5 | 55.8 | |
| OV-AVE2026.06 | 64.9 | 55 | 47.5 | 55.8 | |
| training-free baselinestrategy=training-free2024.11 | 59.8 | 47.3 | 34 | 47 | |
| CLIP & CLAPstrategy=training-free2024.11 | 51.6 | 42.2 | 31.6 | 41.8 | |
| Video-LLaMA2strategy=training-free2024.11 | 48.5 | 38.5 | 29 | 38.6 | |
| AVEstrategy=fine-tuning2024.11 | 44.6 | 33.2 | 24 | 34 | |
| AVE2026.06 | 44.6 | 33.2 | 24 | 34 | |
| MM-Pyramidstrategy=fine-tuning2024.11 | 36.8 | 29 | 23.8 | 29.9 | |
| MM-Pyramid2026.06 | 36.8 | 29 | 23.8 | 29.9 | |
| CMRAstrategy=fine-tuning2024.11 | 36 | 31 | 26.3 | 31.1 | |
| CMRA2026.06 | 36 | 31 | 26.3 | 31.3 | |
| PSPstrategy=fine-tuning2024.11 | 33.7 | 28.2 | 24.2 | 28.7 | |
| PSP2026.06 | 33.7 | 28.2 | 24.2 | 28.7 |