Text-to-Audio Retrieval on Clotho 1K 1.0 (test)
26.9R@1VAST
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| VAST# (audio, text)=27M, Training Regime=internet-scale2023.12 | 26.9 | 53.2 | 66.1 | |
| ONE-PEACE# (audio, text)=2.4M, Training Regime=internet-scale2023.12 | 22.4 | 49 | 62.7 | |
| FuseMix(W&H,B)# (audio, text)=65K, Training Regime=low-data regime2023.12 | 17.6 | 41.1 | 53.5 | |
| VALOR# (audio, text)=6.5M, Training Regime=internet-scale2023.12 | 17.5 | 42.7 | 55.3 | |
| LAION-CLAP# (audio, text)=700K, Training Regime=internet-scale2023.12 | 17.2 | 42.9 | 55.4 | |
| CLAP-HTSAT# (audio, text)=70K, Training Regime=low-data regime2023.12 | 16.8 | 41.1 | 54.1 | |
| FuseMix(W&H,B)# (audio, text)=50K/15K, Training Regime=low-data regime2023.12 | 15.7 | 39.4 | 53.8 | |
| ML-ACT# (audio, text)=50K/15K, Training Regime=low-data regime2023.12 | 14.4 | 36.6 | 49.9 | |
| LAION-CLAP# (audio, text)=65K, Training Regime=low-data regime2023.12 | 12 | 31.6 | 43.9 | |
| MMT# (audio, text)=50K/15K, Training Regime=low-data regime2023.12 | 6.5 | 21.6 | 32.8 |