Text-to-Audio Generation on Clotho (test)
17.23FIDMake-An-Audio
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Make-An-AudioText-cond=T5-Large, Params=563M, Zero-shot evaluation=true2023.01 | 17.23 | 7.02 | — | — | — | |
| Make-An-AudioText-cond=CLAP, Params=332M, Zero-shot evaluation=true2023.01 | 17.38 | 6.98 | — | — | — | |
| Make-An-AudioText-cond=CLIP, Params=576M, Zero-shot evaluation=true2023.01 | 17.55 | 7.09 | — | — | — | |
| Auffusion-FullPretrain=true, Duration(h)=1990, Params=1.1B2024.01 | 17.97 | 2.38 | 1.96 | 11.29 | 55 | |
| Make-An-AudioText-cond=BERT, Params=809M, Zero-shot evaluation=true2023.01 | 18.75 | 7.01 | — | — | — | |
| Auffusion-FullPretrain=false, Duration(h)=1990, Params=1.1B2024.01 | 19.14 | 2.42 | 1.99 | 10.33 | 52.8 | |
| AudioGen-v2-mediumPretrain=false, Duration(h)=6824, Params=1.5B2024.01 | 23.26 | 2.56 | 2.55 | 7.19 | 46.7 | |
| AudioLDM2-largePretrain=true, Duration(h)=29510, Params=1.5B2024.01 | 23.31 | 2.41 | 3 | 8.88 | 49 | |
| AudioLDM2Pretrain=false, Duration(h)=29510, Params=1.1B2024.01 | 23.53 | 2.47 | 3.06 | 9.05 | 48.1 | |
| AudioLDM-L-fullPretrain=false, Duration(h)=9031, Params=975M2024.01 | 24.13 | 2.56 | 3.02 | 7.49 | 51 | |
| DiffsoundText-cond=CLIP, Params=520M, Zero-shot evaluation=true2023.01 | 24.97 | 6.53 | — | — | — | |
| AuffusionPretrain=true, Duration(h)=145, Params=1.1B2024.01 | 25.64 | 2.59 | 2.35 | 9.01 | 48.2 | |
| Tango-FullPretrain=true, Duration(h)=3400, Params=1.3B2024.01 | 25.83 | 2.35 | 3.17 | 6.51 | 50.3 | |
| AudioLDM-S-full-v2Pretrain=false, Duration(h)=9031, Params=421M2024.01 | 26.51 | 2.62 | 3.54 | 6.58 | 49.9 | |
| AuffusionPretrain=false, Duration(h)=145, Params=1.1B2024.01 | 29.01 | 2.66 | 2.67 | 9.46 | 47.6 | |
| RiffusionPretrain=true, Duration(h)=1990, Params=1.1B2024.01 | 31.63 | 2.66 | 6.11 | 6.5 | 47.5 | |
| TangoPretrain=false, Duration(h)=145, Params=1.3B2024.01 | 31.67 | 2.57 | 3.22 | 7.18 | 46.6 |