Table 4 Performance of different frontends with DML improvements on the v1 dataset.
From: Dynamic convolution models for cross-frontend keyword spotting
Network types | Independent | DML-enhanced | DML-improvement | ||||
|---|---|---|---|---|---|---|---|
Net1 | Net2 | Net1 | Net2 | Net1 | Net2 | Net1 | Net2 |
LEAF+ TempConvNet | MFCC+ TempConvNet | 96.2 | 96.37 | 96.3 | 96.8 | 0.1 | 0.43 |
LEAF+ DynTempConvNet | MFCC+ DynTempConvNet | 96.47 | 96.50 | 96.60 | 97 | 0.13 | 0.5 |
