AudioAugmenterBase<T>

Base Classes ยท AiDotNet.Augmentation.Audio

Base class for audio data augmentations.

Inheritance: AugmentationBase<T, Tensor<T>> โ†’ AudioAugmenterBase<T>

Known derived types: AudioNoise<T>, PitchShift<T>, TimeShift<T>, TimeStretch<T>, VolumeChange<T>

For Beginners

Audio augmentation transforms sound data to improve model robustness to variations in recording conditions, speaking styles, and environmental noise. Common techniques include:

  • Time stretching (faster/slower without pitch change)
  • Pitch shifting (higher/lower without speed change)
  • Adding background noise
  • Volume changes
  • Time shifting (moving audio forward/backward)

How It Works

Audio data is typically represented as a 1D waveform tensor or 2D spectrogram.

Properties

Property Summary
int SampleRate { get; set; } Gets or sets the sample rate of the audio data in Hz.

Methods

Method Summary
IDictionary<string, object> GetParameters()