Wolfram Language

Classify Instruments Using Audio Identify Features

All of the high-level machine learning functions such as Classify leverage a variation of the net used in AudioIdentify for feature extraction. This means that the creation of powerful audio classifiers and predictors requires minimal user input. This example collects samples of different musical instrument from the web and classifies them.

Collect samples of guitars, drums and trumpets using WebAudioSearch.

Separate training and testing subsets of the data.

Construct a ClassifierFunction that classifies audio recordings as a guitar, a trumpet or drums.

Use ClassifierMeasurements to show a report of the classification on the test set.

Related Examples

ja