SpeakerVerifier<T>

Models & Types · AiDotNet.Audio.Speaker

Verifies speaker identity by comparing embeddings against enrolled speakers.

Inheritance: NeuralNetworkBase<T> → AudioNeuralNetworkBase<T> → SpeakerRecognitionBase<T> → SpeakerVerifier<T>

Implements: ISpeakerVerifier<T>

For Beginners

Speaker verification is like voice-based password checking:

  1. First, you “enroll” a speaker by recording their voice samples
  2. Later, when someone claims to be that person, you record them and compare
  3. If the voices match closely enough, the identity is verified

Usage (ONNX Mode):

Usage (Native Training Mode):

How It Works

Speaker verification answers the question “Is this the person they claim to be?” by comparing a test utterance against enrolled speaker embeddings.

Constructors

Constructor Summary
SpeakerVerifier() Creates a SpeakerVerifier with default settings for native training mode.
SpeakerVerifier(NeuralNetworkArchitecture<T> architecture, IGradientBasedOptimizer<T, Tensor<T>, Tensor<T>> optimizer, ILossFunction<T> lossFunction, SpeakerVerifierOptions options) Creates a SpeakerVerifier for native training mode.
SpeakerVerifier(NeuralNetworkArchitecture<T> architecture, string embeddingModelPath, SpeakerVerifierOptions options) Creates a SpeakerVerifier for ONNX inference with a pretrained model.

Properties

Property Summary
ISpeakerEmbeddingExtractor<T> EmbeddingExtractor { get; } Gets the underlying speaker embedding extractor.
T DefaultThreshold { get; } Gets the default verification threshold.
bool IsOnnxMode { get; } Gets whether the model is in ONNX inference mode.
double VerificationThreshold { get; } Gets the verification threshold.
int EnrolledCount { get; } Gets the number of enrolled speakers.

Methods

Method Summary
IReadOnlyList<string> GetEnrolledSpeakers() Gets all enrolled speaker IDs.
IdentificationResult Identify(SpeakerEmbedding<T> testEmbedding) Identifies the most likely speaker from enrolled set (legacy API).
ModelMetadata<T> GetModelMetadata() Gets metadata about the model.
SpeakerProfile<T> Enroll(IReadOnlyList<Tensor<T>> enrollmentAudio) Enrolls a speaker by creating a reference embedding from a single audio sample.
SpeakerProfile<T> Enroll(Tensor<T> enrollmentAudio) Enrolls a speaker by creating a reference embedding from a single audio sample.
SpeakerProfile<T> UpdateProfile(SpeakerProfile<T> existingProfile, Tensor<T> newAudio) Updates an existing speaker profile with additional audio.
SpeakerVerificationResult<T> Verify(Tensor<T> audio, Tensor<T> referenceEmbedding) Verifies if a test embedding matches an enrolled speaker (legacy API).
SpeakerVerificationResult<T> Verify(Tensor<T> audio, Tensor<T> referenceEmbedding, T threshold) Verifies if audio matches a reference speaker embedding with custom threshold.
SpeakerVerificationResult<T> VerifyWithReferenceAudio(Tensor<T> audio, Tensor<T> referenceAudio) Verifies if audio matches reference audio of a claimed speaker.
T ComputeScore(Tensor<T> audio, Tensor<T> referenceEmbedding) Computes the verification score between audio and a reference.
T GetThresholdForFAR(double targetFAR) Gets the recommended threshold for a target false accept rate.
Task<SpeakerVerificationResult<T>> VerifyAsync(Tensor<T> audio, Tensor<T> referenceEmbedding, CancellationToken cancellationToken) Verifies if audio matches a reference speaker embedding asynchronously.
Tensor<T> ForwardForTraining(Tensor<T> input)
VerificationResult Verify(string speakerId, SpeakerEmbedding<T> testEmbedding) Verifies if a test embedding matches an enrolled speaker (legacy API).
bool IsEnrolled(string speakerId) Checks if a speaker is enrolled.
bool Unenroll(string speakerId) Removes a speaker’s enrollment.
void Enroll(string speakerId, SpeakerEmbedding<T>[] embeddings) Enrolls a speaker with one or more embeddings (legacy API).