Table 7 Experimental results in the Voxceleb1 development dataset.

From: TDNN achitecture with efficient channel attention and improved residual blocks for accurate speaker recognition

Model

EER

minDCF

ACC (%)

Params (M)

TDNN(C=512)39

0.112

0.806

94.6

3.2

ECAPA-TDNN(C=512)9

0.0654

0.514

98.5

6.9

Res2Net40

0.104

0.779

96.1

5.6

EResNet41

0.0753

0.651

97.2

6.6

ResNetSE42

0.0896

0.715

97.6

7.8

EPCNet-TDNN(C=512)

0.0576

0.423

98.2

13.9

Back to article page