Using Deep Learning to Discover Drugs, Classify Pokémon, Save Zebras, Play Flappy Bird & More
Transmission #2
This week’s newsletter includes using deep learning to: make first impressions, judge a book by its cover, detect ‘ahems’, classify Pokémon and zebra stripes, discover drugs, remove glasses from faces, and play Flappy Bird.
Each week I (@olivercameron) will be sharing the very latest news in deep learning and self-driving cars. To get priority access to these newsletters, please join the mailing list at transmission.ai!
Machine-Vision Algorithm Learning to Judge People by Their Faces
First impressions are notoriously subjective (and flawed), but now machines are being trained to make similar snap judgments based on human-generated data. Read more…
Deep Neural Network Learns to Judge Books by Their Covers
AI-published books may not be too far away! Researchers at Kyushu University in Japan have trained a deep neural network to study book covers and determine their category. Although a great start, the model isn’t quite to human standards just yet. Read more…
20 Weird & Wonderful Datasets for Machine Learning
On a hunt for interesting and high-quality datasets to use for machine learning, I stumbled upon these 20 weird and wonderful sets. Check em’ out!
Deep Learning ‘Ahem’ Detector
Neat! Check out this open source deep convolutional neural network that is trained on transformed audio signals to recognize “ahem” sounds. Potentially very useful for the awkward speakers out there (myself included)! Check out these slides to learn more…
Tesla’s Autopilot 2.0
Jen-Hsun Huang (NVIDIA CEO): “And I think what Tesla has done by launching and having on the road in the very near-future here, a full autonomous driving capability using AI, that has sent a shock wave through the automotive industry. It’s basically five years ahead. Anybody who’s talking about 2021 and that’s just a non-starter anymore.” Read more…
Pokémon, Colors, and Deep Learning
A fun write-up on an attempt to classify 146 Pokémon using a CNN. The most positive outcome of the model showed an accuracy of 96.57%, but there’s still work to do to get reliable results. Read more here…
How Deep Learning Is Helping Save Endangered Zebras
HotSpotter is a set of deep convolutional neural network algorithms that comb through images and identify a zebra by its barcode-like stripes and body shape. Amazing to see deep learning be used for such a good cause! Read more…
Low Data Drug Discovery
The work Vijay Pande and team have been doing recently has been tremendous (see DeepChem and more). This is a must read! “Recent advances in machine learning have made significant contributions to drug discovery. However, the applicability of these techniques has been limited by the requirement for large amounts of training data. In this work, we demonstrate how one-shot learning can be used to significantly lower the amounts of data required to make meaningful predictions in drug discovery applications.” Read more…
Using Deep Learning to Remove Eyeglasses from Faces
An awesome write-up on attempting to automate removal of eyeglasses from a face using deep learning. “Wouldn’t it be great if people could leave their glasses on, and the software automatically removed them? Imagine walking into a retail store and having a virtual mirror remove your glasses and replace them with different products in real-time.” Read more…
Flappy Learning
Remember Flappy Bird, the incredibly frustratingly mobile game? Machine learning has now made easy work of navigating the notorious pipes, and the code is open for the world to see. Try the demo…
That’s it for this week, thanks for reading! If you have any thoughts or questions, I’d love to hear from you in Tweet-form. You can follow and message me at @olivercameron.
