EURASIP Journal on Audio, Speech, and Music Processing

Table 1 Distribution of speech and music events over the 40,906 downloaded balanced train + evaluation segments

From: Exploring convolutional, recurrent, and hybrid deep neural networks for speech and music detection in a large audio dataset

	No-music (%)	Music (%)	Total (%)
No-speech	52.75	20.99	73.74
Speech	19.44	6.83	26.26
Total	72.19	27.81	100

Back to article page