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

Table 4 All-accuracy (%) under various data enhancement methods

From: Deep semantic learning for acoustic scene classification

Mixup

Yes

Yes

Yes

Yes

ImageDtaeGenerator

Yes

No

Yes

Yes

Temporal crop

Yes

Yes

No

Yes

Specaugment

Yes

Yes

Yes

No

All-accuracy(train/test)

80.49/71.26

90.51/69.27

77.56/67.34

85.56/68.32

Back to article page