AudioBuilderExtensions

Helpers & Utilities · AiDotNet

Audio event detection and classification extensions for AiModelBuilder.

For Beginners

Audio event detection identifies sounds in audio recordings or real-time streams. After configuring a model (like BEATs or AudioEventDetector), use these methods to detect sounds, get probabilities, or start real-time monitoring.

How It Works

These methods provide audio-specific operations through the facade pattern. Configure any model that implements IAudioEventDetector via ConfigureModel and then use these methods for audio event detection.

Methods

Method Summary
AudioEventResult<T> DetectAudioEvents(AiModelBuilder<T, Tensor<T>, Tensor<T>> builder, Tensor<T> audio) Detects audio events in the given audio using the configured model’s default threshold.
AudioEventResult<T> DetectAudioEvents(AiModelBuilder<T, Tensor<T>, Tensor<T>> builder, Tensor<T> audio, T threshold) Detects audio events with a custom confidence threshold.
AudioEventResult<T> DetectSpecificAudioEvents(AiModelBuilder<T, Tensor<T>, Tensor<T>> builder, Tensor<T> audio, IReadOnlyList<string> eventTypes) Detects only specific event types, filtering out everything else.
AudioEventResult<T> DetectSpecificAudioEvents(AiModelBuilder<T, Tensor<T>, Tensor<T>> builder, Tensor<T> audio, IReadOnlyList<string> eventTypes, T threshold) Detects specific event types with a custom confidence threshold.
IReadOnlyList<string> GetSupportedAudioEvents(AiModelBuilder<T, Tensor<T>, Tensor<T>> builder) Gets the list of event types the configured audio model can detect.
IStreamingEventDetectionSession<T> StartAudioEventStreaming(AiModelBuilder<T, Tensor<T>, Tensor<T>> builder) Starts a streaming event detection session for real-time audio monitoring.
IStreamingEventDetectionSession<T> StartAudioEventStreaming(AiModelBuilder<T, Tensor<T>, Tensor<T>> builder, int sampleRate, T threshold) Starts a streaming event detection session with custom settings.
Task<AudioEventResult<T>> DetectAudioEventsAsync(AiModelBuilder<T, Tensor<T>, Tensor<T>> builder, Tensor<T> audio, CancellationToken cancellationToken) Detects audio events asynchronously without blocking the calling thread.
Tensor<T> GetAudioEventProbabilities(AiModelBuilder<T, Tensor<T>, Tensor<T>> builder, Tensor<T> audio) Gets raw frame-level event probabilities for all classes without thresholding.
double GetAudioEventTimeResolution(AiModelBuilder<T, Tensor<T>, Tensor<T>> builder) Gets the time resolution of the configured audio event detection model.