Open-Vocabulary Audio-Visual Event Localization on AVEBench (OV seen)
76.6AccuracyAVE
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| AVEstrategy=fine-tuning2024.11 | 76.6 | 63.6 | 56 | 65.4 | |
| MM-Pyramidstrategy=fine-tuning2024.11 | 76.5 | 66.9 | 62.3 | 68.6 | |
| PSPstrategy=fine-tuning2024.11 | 75.4 | 66.8 | 61 | 67.7 | |
| fine-tuning baselinestrategy=fine-tuning2024.11 | 72.5 | 61.8 | 54.5 | 62.9 | |
| CMRAstrategy=fine-tuning2024.11 | 65.2 | 58.8 | 54.3 | 59.4 | |
| training-free baselinestrategy=training-free2024.11 | 57.5 | 45 | 34 | 45.5 | |
| CLIP & CLAPstrategy=training-free2024.11 | 51.4 | 41.4 | 31.9 | 41.6 | |
| Video-LLaMA2strategy=training-free2024.11 | 50.1 | 40.6 | 32 | 40.9 |