[News] Facebook's Real-Time TTS system runs on CPUs only!

Channel:
Switzerland Yannic Kilcher
Subscribers:
291,000
Published on ● Video Link: https://www.youtube.com/watch?v=XvDzZwoQFcU


Duration: 21:17
6,022 views
230

Facebook AI's new Text-To-Speech system is able to create 1 second of speech in as little as 500ms, making it real-time. What's even more impressive is the fact that this does not require a rack of GPUs, but runs on merely 4 CPUs.

OUTLINE:
0:00 - Intro
1:00 - Problem Formulation
3:20 - System Explanation
15:00 - Speeding up the computation

https://ai.facebook.com/blog/a-highly-efficient-real-time-text-to-speech-system-deployed-on-cpus/

Links:
YouTube: https://www.youtube.com/c/yannickilcher
Twitter: https://twitter.com/ykilcher
BitChute: https://www.bitchute.com/channel/yannic-kilcher
Minds: https://www.minds.com/ykilcher



Other Videos By Yannic Kilcher


2020-05-26A critical analysis of self-supervision, or what we can learn from a single image (Paper Explained)
2020-05-25Deep image reconstruction from human brain activity (Paper Explained)
2020-05-24Regularizing Trajectory Optimization with Denoising Autoencoders (Paper Explained)
2020-05-23[News] The NeurIPS Broader Impact Statement
2020-05-22When BERT Plays the Lottery, All Tickets Are Winning (Paper Explained)
2020-05-21[News] OpenAI Model Generates Python Code
2020-05-20Investigating Human Priors for Playing Video Games (Paper & Demo)
2020-05-19iMAML: Meta-Learning with Implicit Gradients (Paper Explained)
2020-05-18[Code] PyTorch sentiment classifier from scratch with Huggingface NLP Library (Full Tutorial)
2020-05-17Planning to Explore via Self-Supervised World Models (Paper Explained)
2020-05-16[News] Facebook's Real-Time TTS system runs on CPUs only!
2020-05-15Weight Standardization (Paper Explained)
2020-05-14[Trash] Automated Inference on Criminality using Face Images
2020-05-13Faster Neural Network Training with Data Echoing (Paper Explained)
2020-05-12Group Normalization (Paper Explained)
2020-05-11Concept Learning with Energy-Based Models (Paper Explained)
2020-05-10[News] Google’s medical AI was super accurate in a lab. Real life was a different story.
2020-05-09Big Transfer (BiT): General Visual Representation Learning (Paper Explained)
2020-05-08Divide-and-Conquer Monte Carlo Tree Search For Goal-Directed Planning (Paper Explained)
2020-05-07WHO ARE YOU? 10k Subscribers Special (w/ Channel Analytics)
2020-05-06Reinforcement Learning with Augmented Data (Paper Explained)


Tags:
deep learning
machine learning
arxiv
explained
neural networks
ai
artificial intelligence
paper
facebook
fair
tts
text-to-speech
real-time
wavenet
rnn
spectrogram
mel
frequency
vocoder
linguistic
features
speaker
soft
style
tone
phonemes
neural
recurrent
human
assistant