Skip to main content


Fig. 5 | EURASIP Journal on Audio, Speech, and Music Processing

Fig. 5

From: Biomimetic spectro-temporal features for music instrument recognition in isolated notes and solo phrases

Fig. 5

Fisher discriminant on combined solo and isolated notes for different feature sets. Average log F-ratio is depicted for each feature set; using rate-scale-frequency, scale-frequency, rate-frequency, and scale-rate. The analysis is performed on combined solo and isolated notes and tests the separability of this combined set to identify each instrument class. The bar plot shows the mean log F-ratio, error bars indicate standard deviation

Back to article page