AcousticModelBase<T>

Base Classes · AiDotNet.TextToSpeech

Base class for acoustic TTS models that generate mel-spectrograms from text.

Inheritance: NeuralNetworkBase<T> → TtsModelBase<T> → AcousticModelBase<T>

Implements: IAcousticModel<T>

How It Works

Acoustic models form the first stage of a two-stage TTS pipeline: Text -> [Acoustic Model] -> Mel-Spectrogram -> [Vocoder] -> Waveform.

Subclasses include Tacotron, Tacotron2, FastSpeech, FastSpeech2, GlowTTS, GradTTS, and other models that produce mel-spectrograms requiring a separate vocoder for audio output.

Properties

Property Summary
int FftSize { get; } Gets the FFT window size.
int MaxTextLength { get; } Gets the maximum text length.

Methods

Method Summary
Tensor<T> Synthesize(string text)
Tensor<T> TextToMel(string text)