Two Minute Papers explores the paper “ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness ” The source documents and code are available at the following links:
YouTuber carykh explores GANs that create customized faces of imaginary celebrities.
Using neural networks can you generate generate new, never before seen Garfield comics? CodeParade explains how it works.
Here’s a great video explaining Convolutional Neural Networks (CNNs), a type of neural network used in computer vision scenarios.
This Kaggle explains the inner workings of transfer learning, a fascinating application of neural networks.
Edwin shared a tweet that encapsulates all the common AI terms
Whatever you need to know about #AI#ML #DL #ComputerVision@MikeQuindazzi @evankirstel @psb_dc @diioannid @SpirosMargaris @helene_wpli @Paula_Piccard @mclynd @andi_staub @ipfconline @LouisSerge @jerome_joffre @ahier #edmuke @Ym78200 @3itcom pic.twitter.com/GMvcm6Mt22 @jblefevre60
— Edwin (@edmuke) November 12, 2018
Another great video from DeepLearning.TV.
At the recent Azure AI Fest, I had mentioned the work being done to train self-driving car AIs using video games like Grand Theft Auto V.
And, yes, Virginia, a kid on his own PC at home can compete with a multi-million dollar company at the same task. For reference, see Facebook, Microsoft, and Apple. True story.
How did OpenAI’s team of 5 neural networks manage to beat some of the world’s best DOTA 2 players?
Watch this video by Siraj Raval to learn how the team did it and what it means for AI research.