Pushing the frontier of neural text to speech

In the popular field of text to speech, the goal is to transform the written or printed word into speech that is natural and intelligible. Today, the technology is being used in products and services to help people who are blind or have low vision consume digital content, power personal digital assistants that sound more realistic, and make it easier to do two things at once, such as listening to an article online while washing dishes, among other applications. Although the quality of synthesized speech has gotten better thanks to neural network-based end-to-end TTS, advancing neural TTS and allowing it to be more easily integrated into product development and deployment requires overcoming a variety of remaining challenges.

In this webinar, Senior Researcher Xu Tan will talk about these challenges, specifically the high computational cost and slow inference speed in online serving; word skipping and repeating issues, poor voice quality, and lack of voice controllability; the large amounts of training data needed for improved voice synthesis; and the practical challenges in TTS voice adaptation. He’ll introduce his team’s work addressing these challenges—including fast TTS, end-to-end TTS, low-resource TTS, and adaptive TTS—as well as discuss other critical questions and opportunities to pursue in the space.

Together, you’ll explore:

  • An overview of text to speech, including its evolution
  • The important challenges in neural text to speech and how to address them with dedicated research
  • How to factor product development into your research

Resource list:

*This on-demand webinar features a previously recorded Q&A session and open captioning.

This webinar originally aired on May 20, 2021

Explore more Microsoft Research webinars: https://aka.ms/msrwebinars

Date:
Speakers:
Xu Tan
Affiliation:
Microsoft Research