MfccExtractor<T>

Models & Types · AiDotNet.Audio.Features

Extracts Mel-Frequency Cepstral Coefficients (MFCCs) from audio signals.

Inheritance: AudioFeatureExtractorBase<T> → MfccExtractor<T>

For Beginners

MFCCs capture the “shape” of the audio’s frequency content, similar to how humans perceive sound. The process:

  • Compute the Mel spectrogram (power spectrum on perceptual scale)
  • Take the log (matches human loudness perception)
  • Apply DCT (decorrelates and compresses the information)
  • Keep only the first N coefficients (typically 13-40)

Why MFCCs work well for speech:

  • They capture formant frequencies (vocal tract resonances)
  • They’re robust to background noise
  • They compress audio information efficiently

Usage:

How It Works

MFCCs are a compact representation of the spectral envelope of an audio signal. They are widely used in speech recognition, speaker identification, and music analysis.

Constructors

Constructor Summary
MfccExtractor(MfccOptions options) Initializes a new MFCC extractor.

Properties

Property Summary
int FeatureDimension { get; }
string Name { get; }

Methods

Method Summary
Tensor<T> Extract(Tensor<T> audio)