Skip to main content

Advertisement

Table 11 Mean error rates of instrument recognition methods (italics indicates the best result for the overall performance)

From: A computational study of auditory models in music recognition tasks for normal-hearing and hearing-impaired listeners

  Polyphonic Monophonic
Method Cello vs. all Clarinet vs. all Trumpet vs. all Overall Overall
AM features, RF 0.012 0.017 0.029 0.019 0.002
AM features, SVM 0.007 0.007 0.014 0.011 0.001
Standard features, RF 0.044 0.034 0.052 0.063 0.000
Standard features, SVM 0.025 0.019 0.054 0.035 0.002