Speech Transcription and Synthesis

Use a pretrained model or third-party APIs for text-to-speech and speech-to-text

Audio Toolbox™ provides examples for small-vocabulary recognition and sound synthesis. Use the wav2vec 2.0 pretrained network to perform general speech-to-text transcription with speech2text. You can download Audio Toolbox extended functionality from File Exchange for text-to-speech and speech-to-text through interfaces to popular third-party APIs. Supported APIs include Google^® Speech, IBM^® Watson Speech, and Microsoft^® Azure Speech.

You can interact with speech-to-text functionality graphically in the Signal Labeler app to quickly label regions of speech.

Apps

Signal Labeler

Label signal attributes, regions, and points of interest, and extract features

Functions

`speech2text`	Transcribe speech signal to text
`text2speech`	Synthesize speech from text
`speechClient`	Interface with pretrained model or third-party speech service

Topics

Label Spoken Words in Audio Signals
Use Signal Labeler to label spoken words in an audio signal.

Featured Examples

Train Speech Command Recognition Model Using Deep Learning

Train a deep learning model that detects the presence of speech commands in audio.

Open Live Script

Keyword Spotting in Noise Using MFCC and LSTM Networks

Identify a keyword in noisy speech using a deep learning network. In particular, the example uses a Bidirectional Long Short-Term Memory (BiLSTM) network and mel frequency cepstral coefficients (MFCC).

Open Live Script

Spoken Digit Recognition with Wavelet Scattering and Deep Learning

Classify spoken digits using both machine and deep learning techniques. In the example, you perform classification using wavelet time scattering with a support vector machine (SVM) and with a long short-term memory (LSTM) network. You also apply Bayesian optimization to determine suitable hyperparameters to improve the accuracy of the LSTM network. In addition, the example illustrates an approach using a deep convolutional neural network (CNN) and mel-frequency spectrograms.

Open Live Script

Train Generative Adversarial Network (GAN) for Sound Synthesis

Train and use a generative adversarial network (GAN) to generate sounds.

Open Live Script

LPC Analysis and Synthesis of Speech

Use the Levinson-Durbin and Time-Varying Lattice Filter blocks for low-bandwidth transmission of speech using linear predictive coding.

Open Model