vad

Documentation

Index

• func FramesBinaryThresholdTuned(ctx context.Context, pcm []int16, mode Mode, threshold float64, ...) []float64
• type Mode
• type Tuning

Functions

func FramesBinaryThresholdTuned

func FramesBinaryThresholdTuned(ctx context.Context, pcm []int16, mode Mode, threshold float64, overhangFrames int, adaptScale float64, tuning Tuning) []float64

FramesBinaryThresholdTuned returns binary {-1, +1} with custom threshold, overhang, adaptation, and GMM tuning parameters. PCM must be 8kHz mono s16le.

Types

type Mode

type Mode int

Mode represents the VAD aggressiveness level (WebRTC modes 0-3).

const (
	ModeQuality        Mode = 0 // Most permissive, fewest false negatives
	ModeLowBitrate     Mode = 1 // Low bitrate
	ModeAggressive     Mode = 2 // Aggressive
	ModeVeryAggressive Mode = 3 // Most restrictive, fewest false positives
)

Mode constants for the supported WebRTC VAD aggressiveness levels.

type Tuning

type Tuning struct {
	// contains filtered or unexported fields
}

Tuning holds optional GMM tuning parameters.