eBook
End-to-End Speech AI Pipelines
An in-depth explainer of ASR and TTS, the two main components of Speech AI.
What’s Included in This eBook?
Automatic Speech Recognition (ASR) and Text-To-Speech (TTS) are the two most essential Speech AI technologies. Each of these technological pipelines includes multiple stages, such as data preprocessing, deep learning models, and post-processing. This eBook details what occurs in each of their individual components and how to evaluate the performance of these technologies.
What Is Automatic Speech Recognition?
ASR, also known as speech-to-text, is the process of automatically converting spoken audio into written form.
What Is Text-to-Speech?
TTS, also known as speech synthesis, takes text as an input and generates a human-like synthesized voice.
How Do I Evaluate ASR and TTS?
Metrics, such as word error rate (WER) and mean opinion score (MOS), are used to assess the performance of ASR and TTS pipelines, respectively.
Register to Download