Learn Reinforcement Learning for free with Google, Neural Network vs. Neural Network at AI World Cup, and more Machine Learning Resources!
Sep 3, 2018 · 3 min read
This was yet another week headlined by Google’s releases. By open sourcing TensorFlow code and dataset for reinforcement learning, they have spiked interest in this field and paved the way for everyone to learn and get involved.
Other highlights from the past week: Another Google release — this time a Text-to-Speech offering for everyone, turn your handwritten notes into HTML code in seconds, more enthralling resources and articles below!
Subscribe here to get daily AVBytes articles delivered every morning, right to your inbox!

- Google AI Open Sources TensorFlow Code & Dataset for Reinforcement Learning: Want to understand reinforcement learning? You won’t get a better chance — Google AI has open sourced their entire dataset and a TensorFlow-based framework to help you get started in this field. Read more inside and download the code and dataset now!

- Sketch2Code Transforms your Handwritten Notes into Working Code within Seconds — and it’s Open Source!: A very useful offering from Microsoft — Sketch2Code takes your handwritten notes, and converts them into working code within seconds, all thanks to computer vision! It’s open source so you can try it out right now.

- A Neural Network Algorithm Analyzes Eye Blinking to Detect Fake Videos (with 95% Accuracy): Fake videos are becoming better at fooling existing techniques. So here’s an algorithm that detects these fake videos with an incredible 95% accuracy! The concept is unique and worth exploring more.

- Google Cloud’s Machine Learning Powered Text-to-Speech is Available for Everyone!: It’s easier than ever to combine NLP with audio processing, thanks to Google Cloud’s Text-to-Speech API! 17 new WaveNet voices have been added, and this article also contains resources for the TensorFlow implementation of this model. Download the code today!
- AI World Cup is a Fully Simulated Football Tournament Powered by Neural Networks: A reinforcement learning algorithm just won the recently concluded ‘AI World Cup’! It was as intriguing as it sounds, with 5-a-side virtual players powered by neural networks taking the field to win the ultimate prize. This article covers the algorithm and a superb video demo!

The above AVBytes were published from 27th August to 2nd September, 2018.
