Interesting food for thought from the Director of AI Research at Facebook and Professor at NYU.

Watch NYC being turned into a moving painting. NYC FLOW is an exploration of video processing technics introduced by neural-style code. 

In this post, the OpenAI team lists several problem areas likely to be important both for advancing AI and for its long-run impact on society. Applications are open.

Our kids don’t need to be taught in order to learn. Research shows that “kids play with toys that "give them the most information about how the world works.” I love the resemble to Reinforcement Learning.

Google’s Bubble Zoom uses Machine Learning to zoom in on text bubbles in comics with one touch.

This post explains the problem of Inverse Reinforcement Learning - Given the optimal expert policy, how can we determine the underlying reward structure?

Learn how to use Torch with Noise Contrastive Estimation (NCE) to train a multi-GPU recurrent neural network language model (RNNLM) on the Google billion words (GBW) dataset.

Making Neural Machine Translation work can be tricky - This repo goes over lots of practical tips, using the lamtram toolkit.

DeepHeart is a neural network designed for the 2016 Physionet Challenge in predicting cardiac abnormalities from phonocardiogram (PCG) data

Directory of tutorials and open-source code repositories for working with Keras, the Python deep learning library

A new benchmark task to recognize one million celebrities from their face images, by using all the possibly collected face images of this individual on the web as training data. By Yandong Guo, Lei Zhang, Yuxiao Hu, Xiaodong He, Jianfeng Gao (Microsoft)

An approach to training neural networks to generate sequences using actor-critic methods from reinforcement learning (RL). By Dzmitry Bahdanau, Philemon Brakel, Kelvin Xu, Anirudh Goyal, Ryan Lowe, Joelle Pineau, Aaron Courville, Yoshua Bengio (UMontreal)

A weakly-supervised framework for action labeling in video, where only the order of occurring actions is required during training time. By De-An Huang, Li Fei-Fei, Juan Carlos Niebles (Stanford)