awesome project: turning youtube videos of soccer games into augmented reality scenes on a tabletop https://goo.gl/18ybiH @ArtificialIntelligenceArticles

OpenAI Five !

A team of five neural networks, OpenAI Five, has started to defeat amateur human teams (including a semi-pro team) at Dota 2 https://goo.gl/CjEq68 @ArtificialIntelligenceArticles

Super fun dataset and challenge : Fashion-Gen: The Generative Fashion Dataset and Challenge

More details about the Fashion-Gen dataset (Text to Image generation) and the challenge:

https://arxiv.org/abs/1806.08317

Participate in the challenge https://fashion-gen.com/ to win 2000$ and submit your paper to

https://sites.google.com/view/eccvfashion/

https://t.iss.one/ArtificialIntelligenceArticles

NIPS 2018: AI for Prosthetics Challenge

Reinforcement learning with musculoskeletal models

Objective : to bring deep reinforcement learning to solve problems in medicine

https://www.crowdai.org/challenges/nips-2018-ai-for-prosthetics-challenge

https://t.iss.one/ArtificialIntelligenceArticles

Meet the Godfather of AI(Geoffrey Hinton )

"It was just obvious to me that it was the right way to go. The brain's a big neural network. And so, it has to be that stuff like this can work, because it works in our brains. There's just never any doubt about that." - Geoffrey Hinton


A new interview with Geoffrey Hinton

video :

https://www.youtube.com/watch?v=l9RWTMNnvi4&feature=youtu.be

https://t.iss.one/ArtificialIntelligenceArticles

Turning Fortnite into PUBG with Deep Learning (CycleGAN)
https://towardsdatascience.com/turning-fortnite-into-pubg-with-deep-learning-cyclegan-2f9d339dcdb0 @ArtificialIntelligenceArticles

In a new paper, Google Research Scientists by Carl Vondrick, Abhinav Shrivastava, Alireza Fathi, Sergio Guadarrama, and Kevin Murphy introduce a network that learns to colorize videos, automatically tracking objects without supervision and without requiring any labeled training data.

https://ai.googleblog.com/2018/06/self-supervised-tracking-via-video.html

Paper: https://arxiv.org/abs/1806.09594


https://t.iss.one/ArtificialIntelligenceArticles

Top 20 DeepLearning Papers, 2018 Edition https://goo.gl/aU19R3 @ArtificialIntelligenceArticles

Evaluating Feature Importance Estimates

Important work from @sarahookr, @doomie, @piekindermans, and @_beenkim on quantifying the extent to which various saliency methods *actually* find relevant portions of images.


https://arxiv.org/abs/1806.10758

https://t.iss.one/ArtificialIntelligenceArticles

Some ideas on how to train neural networks in a somewhat more biologically plausible way compared to end-to-end training with the backpropagation algorithm. Unsupervised Learning by Competing Hidden Units,
https://arxiv.org/abs/1806.10181

@ArtificialIntelligenceArticles

Scalable Deep Reinforcement Learning for Robotic Manipulation

https://ai.googleblog.com/2018/06/scalable-deep-reinforcement-learning.html

https://t.iss.one/ArtificialIntelligenceArticles

Using Deep Q-Learning in FIFA 18 to perfect the art of free-kicks
https://goo.gl/yzQ9bK @ArtificialIntelligenceArticles

Slides for most talks at Good Citizen at CVPR workshop are up https://goo.gl/qXzezt or https://goo.gl/8Q98vR Lots of useful advice and experience for writing and reviewing papers, how to do good research and evaluation, talks, how to organise your time #CVPR2018

https://t.iss.one/ArtificialIntelligenceArticles

Adversarial Reprogramming of Neural Networks. Cool new preprint from Ian Goodfellow ,.... https://arxiv.org/abs/1806.11146 @ArtificialIntelligenceArticles

Beyond Winning and Losing: Modeling Human Motivations and Behaviors (in World of Warcraft) Using Inverse Reinforcement Learning

https://arxiv.org/abs/1807.00366 @ArtificialIntelligenceArticles

The 10 coolest papers from CVPR 2018
https://towardsdatascience.com/the-10-coolest-papers-from-cvpr-2018-11cb48585a49 @ArtificialIntelligenceArticles

Accurate Uncertainties for Deep Learning Using Calibrated Regression https://arxiv.org/abs/1807.00263 @ArtificialIntelligenceArticles

Human-level performance in first-person multiplayer games with population-based deep reinforcement learning"

Agents achieve human-level performance in a complex first-person multiplayer game, and can even collaborate with human teammates!

Jaderberg et al. Blog : https://deepmind.com/blog/capture-the-flag/

@ArtificialIntelligenceArticles

new paper from Geoffrey Hinton #ACL2018 https://goo.gl/TehHaz @ArtificialIntelligenceArticles

The proceedings of #icml2018 are now online
#paper
https://goo.gl/e5y2Yk @ArtificialIntelligenceArticles

Best Paper Awards ICML 2018
Paper: https://arxiv.org/abs/1802.00420
Github: https://github.com/anishathalye/obfuscated-gradients @ArtificialIntelligenceArticles