Table 9 The impact of pre training on network performance.
From: Dense dynamic convolutional network for Bel canto vocal technique assessment
ID | ImageNet | Audioset | Top-1_Acc(%) |
|---|---|---|---|
I | ✖ | ✖ | 90.11 |
II | ✖ | ✔ | 91.23(+1.12) |
III | ✔ | ✖ | 90.56(+0.45) |
IV | ✔ | ✔ | 91.88(+1.77) |
