EnergyBasedVad<T>
Models & Types · AiDotNet.Audio.VoiceActivity
Simple energy-based voice activity detector (algorithmic, no neural network).
Inheritance: VoiceActivityDetectorBase<T> → EnergyBasedVad<T>
For Beginners
This is the simplest type of VAD:
Basic idea: Speech is louder than silence!
- Compute the “energy” (sum of squared samples) for each frame
- If energy exceeds a threshold, it’s probably speech
Enhanced features used here:
- Energy: How loud is the signal?
- Zero-crossings: How often does the signal cross zero?
- Speech: Medium zero-crossings (voiced sounds)
- Noise: High zero-crossings (random noise)
- Spectral flatness: Is it tonal or noisy?
- Speech: Low flatness (has harmonic structure)
- Noise: High flatness (random spectrum)
Pros:
- Very fast (no neural network)
- Low latency
- Works well in quiet environments
Cons:
- Struggles with background noise
- May trigger on loud non-speech sounds
- Requires threshold tuning for different environments
For better noise robustness, use neural network-based VAD like SileroVad.
How It Works
This is a basic VAD that detects speech based on signal energy (loudness). It combines multiple features for more robust detection:
- Short-time energy
- Zero-crossing rate
- Spectral flatness
Constructors
| Constructor | Summary |
|---|---|
EnergyBasedVad(int sampleRate, int frameSize, double threshold, double energyWeight, double zcrWeight, double flatnessWeight, bool adaptiveThreshold, int minSpeechDurationMs, int minSilenceDurationMs) |
Creates an energy-based VAD with default parameters. |
Methods
| Method | Summary |
|---|---|
void ResetState() |
Resets the VAD state including adaptive thresholds. |