EURASIP Journal on Audio, Speech, and Music Processing

Table 11 Accuracy (Acc.) and precision (Prec), recall (Rec) and F1 score (F1) per class and on average for the Albayzín 2010 test data evaluated using the best performing system presented in this paper

From: Multiclass audio segmentation based on recurrent neural networks for broadcast domain data

Class	Prec	Rec	F1
mu	0.88	0.89	0.88
sm	0.93	0.84	0.89
sn	0.85	0.84	0.84
sp	0.83	0.88	0.85
Avg	0.87	0.86	0.87
Acc.			0.85

Back to article page