Audio analysis is achieved by visually or programmatically inspecting local and global features extracted from the signal. For audio signals, analysis may happen in either time or frequency, or both. The Wolfram Language provides a large collection of functions, from low-level spectral analysis to high-level functions for classifying audio signals or recognizing speech.
AudioPlot — waveform plot of audio
Spectrogram — spectrogram or time-frequency plot of audio
Periodogram — power spectrum plot of audio
Cepstrogram — power cepstra plot of audio
AudioDistance — compute a distance measure between two audio objects
AudioBlockMap — apply a function to audio partitions
AudioLoudness — compute different loudness standards of an audio signal
AudioIntervals ▪ AudioMeasurements ▪ AudioLocalMeasurements
ShortTimeFourier — compute short-time Fourier transform (STFT)
Fourier ▪ PeriodogramArray ▪ SpectrogramArray ▪ CepstrogramArray ▪ CepstrumArray ▪ InverseShortTimeFourier ▪ InverseSpectrogram
Understanding Audio Signals
AudioIdentify — attempt to identify what an audio signal is a recording of
PitchRecognize ▪ AudioInstanceQ
Understanding Speech »
SpeechRecognize — convert a spoken audio signal to text
SpeechCases ▪ SpeechInterpreter ▪ ...
AudioAnnotate — annotate an audio object
AudioAnnotationLookup ▪ AnnotationDelete ▪ AnnotationRules
Machine Learning »
Classify, Predict — create and apply classifiers or predictors to audio signals
Nearest ▪ FeatureNearest ▪ FeatureSpacePlot ▪ FindClusters ▪ ...
NetEncoder ▪ NetChain ▪ NetGraph ▪ ...