Extract a Specific Feature

A powerful tool for comparing and visualizing audio recordings is high-level feature extraction. This example shows how to extract specific features from an audio signal.

Collect all the Audio objects in ExampleData.

Compute the "MFCC" features for all of the signals using FeatureExtract.

Combine two feature extractors to produce a fixed-length vector instead of a variable-length tensor and plot the result.

Create a NearestFunction using a user-defined feature extractor.

Find the nearest audio sample.