Here’s a great speech by Ed Mylett that is sure to not only motivate you, but also change your perspective.
I mentioned this on my livestream earlier today and thought it was worth sharing.
That and Jocko Willink is always worth a listen.
Text-to-speech engines are usually multi-stage pipelines that transform the signal into many intermediate representations and require supervision at each step.
When trying to train TTS end-to-end, the alignment problem arises: Which text corresponds to which piece of sound?
This paper uses an alignment module to tackle this problem and produces astonishingly good sound.
Paper: https://arxiv.org/abs/2006.03575
Website: https://deepmind.com/research/publications/End-to-End-Adversarial-Text-to-Speech
Content index:
- 0:00 – Intro & Overview
- 1:55 – Problems with Text-to-Speech
- 3:55 – Adversarial Training
- 5:20 – End-to-End Training
- 7:20 – Discriminator Architecture
- 10:40 – Generator Architecture
- 12:20 – The Alignment Problem
- 14:40 – Aligner Architecture
- 24:00 – Spectrogram Prediction Loss
- 32:30 – Dynamic Time Warping
- 38:30 – Conclusion
Noelle shares this demo from the VOICE Summit showing off Custom Speech and Custom Language Pre-built AI Models
Watch Panos Periorelles, PM on Cognitive Services team, to learn about the latest advancements in using speech recognition and speech synthesis including how to create your own custom model.