Table of Contents

Class AudioAugmentationSettings

Namespace
AiDotNet.Augmentation
Assembly
AiDotNet.dll

Audio-specific augmentation settings with industry-standard defaults.

public class AudioAugmentationSettings
Inheritance
AudioAugmentationSettings
Inherited Members

Remarks

For Beginners: These settings control how audio data is augmented. Defaults are based on best practices from audiomentations and torchaudio.

Properties

EnableNoise

Gets or sets whether background noise is enabled.

public bool EnableNoise { get; set; }

Property Value

bool

Remarks

Default: true

Adds random noise to simulate different recording conditions.

EnablePitchShift

Gets or sets whether pitch shifting is enabled.

public bool EnablePitchShift { get; set; }

Property Value

bool

Remarks

Default: true

Changes the pitch without changing tempo. Good for speech/music tasks.

EnableTimeShift

Gets or sets whether time shift is enabled.

public bool EnableTimeShift { get; set; }

Property Value

bool

Remarks

Default: true

Shifts the audio forward or backward in time.

EnableTimeStretch

Gets or sets whether time stretching is enabled.

public bool EnableTimeStretch { get; set; }

Property Value

bool

Remarks

Default: true

Changes tempo without changing pitch.

EnableVolumeChange

Gets or sets whether volume change is enabled.

public bool EnableVolumeChange { get; set; }

Property Value

bool

Remarks

Default: true

Randomly adjusts volume to simulate different recording levels.

MaxTimeShift

Gets or sets the maximum time shift as a fraction of audio length.

public double MaxTimeShift { get; set; }

Property Value

double

Remarks

Default: 0.1 (10% of audio length)

MaxTimeStretch

Gets or sets the maximum time stretch factor.

public double MaxTimeStretch { get; set; }

Property Value

double

Remarks

Default: 1.2 (120% speed, faster)

MinTimeStretch

Gets or sets the minimum time stretch factor.

public double MinTimeStretch { get; set; }

Property Value

double

Remarks

Default: 0.8 (80% speed, slower)

NoiseSNR

Gets or sets the signal-to-noise ratio in decibels.

public double NoiseSNR { get; set; }

Property Value

double

Remarks

Default: 20.0 (dB)

Higher values mean less noise relative to signal.

PitchShiftRange

Gets or sets the pitch shift range in semitones.

public double PitchShiftRange { get; set; }

Property Value

double

Remarks

Default: 2.0 (+/- 2 semitones)

VolumeChangeRange

Gets or sets the volume change range in decibels.

public double VolumeChangeRange { get; set; }

Property Value

double

Remarks

Default: 6.0 (+/- 6 dB)

Methods

GetConfiguration()

Gets the configuration as a dictionary.

public IDictionary<string, object> GetConfiguration()

Returns

IDictionary<string, object>