Artwork-to-Music Generation on Artwork-to-Music (Evaluation)
13FADArt2Mus
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Art2MusEmb. Model=ImageBind, Train Size (hrs)=≈2.6, Tok/Img=162026.02 | 13 | 0.003 | 0.0467 | 0.2037 | |
| AudioLDM 2Train Size (hrs)=≈1,150,0002026.02 | 14 | 0.0043 | 0.0982 | 0.1741 | |
| Art2MusEmb. Model=CLIP, Tok/Img=62026.02 | 15 | 0.0037 | 0.0417 | 0.1672 | |
| AudioLDM 2-MusicTrain Size (hrs)=≈665,0002026.02 | 15 | 0.0845 | 0.1868 | 0.1868 | |
| Mozart’s TouchTrain Size (hrs)=≈20,0002026.02 | 19 | 0.0024 | 0.0835 | 0.2163 |