Skip to main content
EURASIP Journal on Audio, Speech, and Music Processing

Table 3 Comparison results of several self-attentions

From: Channel and temporal-frequency attention UNet for monaural speech enhancement

 

Times(s)

MACs

Para.

VSA

7.015

317.482 M

3.936 K

ISA [19]

0.064

495.581 M

6.144 K

ASA [20]

0.192

152.933 M

1.752 K

ours.

0.062

400.079 M

4.960 K

Back to article page