Text-to-Sound Generation on Clotho zero-shot (test)
2.13FADMake-an-Audio 2
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Make-an-Audio 2Training Data (Hours)=8.7k2023.10 | 2.13 | 2.49 | 61.52 | 69.9 | |
| AudioGenTraining Data (Hours)=4k2023.10 | 2.55 | 2.5 | 63.84 | 72.12 | |
| UniAudioTraining Data (Hours)=7k2023.10 | 3.12 | 2.57 | 61.9 | 66.1 | |
| TangoTraining Data (Hours)=3.3k2023.10 | 3.61 | 2.59 | 66.2 | 68.57 | |
| AudioLDMTraining Data (Hours)=9k2023.10 | 4.93 | 2.6 | 60.95 | 65.7 | |
| DiffsoundTraining Data (Hours)=2k2023.10 | 7.8 | 6.53 | — | — | |
| Reference2023.10 | — | — | 70.47 | 78.84 |