EURASIP Journal on Audio, Speech, and Music Processing

Table 6 Balanced classification errors for recognition of instrument groups. CNN-AlexNet: adaptation after [63]; CNN-Han: adaptation after [64, 65]; CNN-VGG16: adaptation after [66]. Best values are marked with bold font\(^{a}\)

From: AAM: a dataset of Artificial Audio Multitracks for diverse music information retrieval tasks

Instr.	Random	CNN-	CNN-	CNN-
Group	Forest	Han	AlexNet	VGG16
Bass	0.192	0.017	0.021	0.016
Brass	0.270	0.043	0.094	0.035
Drums	0.159	0.036	0.028	0.035
Guitar	0.346	0.106	0.169	0.095
Organ	0.101	0.002	0.004	0.001
Piano	0.388	0.099	0.162	0.094
Pipe	0.418	0.087	0.145	0.087
Reed	0.389	0.106	0.172	0.117
Strings	0.168	0.030	0.051	0.029

\(^{a}\)CNN-Han is better than CNN-VGG16 for pipe at the 6th position after the fix point

Back to article page