Deep Learning Weekly | Issue #3 facebook open sources fastText, colorize b/w videos, emoji embeddings & mollified neural networks

Bringing you everything new and exciting in the world of
 deep learning from academia to the grubby depth
 of industry every week right to your inbox. Free.

Howdy folks,welcome to our second issue. This week we are regaling you with a veritable cornucopia of
August 23 · Issue #3 · View online
Deep Learning Weekly
Howdy folks,

welcome to our second issue. This week we are regaling you with a veritable cornucopia of interesting developments in deep learning from emoji embeddings over facebooks newly open sourced fastText library to what to do when your gradient descent doesn’t converge.

If you like this issue follow us on twitter, tweet about this issue, tell your colleagues, your spouse, the people in front of you in line at the grocery store… really we are grateful for any word that gets out.

Have a great week.

Correcting Intel’s Deep Learning Benchmark Mistakes | NVIDIA Blog
Nuance brings deep learning tech to its Dragon speech recognition | TechCrunch
Dango - Emoji & Deep Learning
Making Kaggle the Home of Open Data | No Free Hunch
Excire Search for Lightroom Helps You Find Photos with AI
Facebook’s Artificial Intelligence Research lab releases open source fastText on GitHub | TechCrunch
End-to-End Deep Learning for Self-Driving Cars | Parallel Forall
JIT native code generation for TensorFlow computation graphs using Python and LLVM | Terra Incognita
You also get a neat graphic illustrating the RNN they used to embed the emojis, Emoji2Vec anyone?
7 Years - Neural Network Research (Deep Learning) - Coloring a B&W Videoclip - YouTube
Landscape of Deep Learning Frameworks — Medium
Interviews and Q&As
RE•WORK | Blog - Is Humanising Chatbots the Next Inevitable Improvement?
Yann LeCun's answer to What are your recommendations for self-studying machine learning? - Quora
Libraries & Code
GitHub -facebookresearch/fastText: Library for fast text representation and classification.
GitHub - jisaacso/DeepHeart: Neural networks for monitoring cardiac data
GitHub - jnhwkim/ddx: Deep Learning Dashboard
GitHub - rhiever/tpot: A Python tool that automatically creates and optimizes machine learning pipelines using genetic programming.
Papers & Publications
Mollifying Networks
Stochastic Gradient Descent with Restarts
Full Resolution Image Compression with Recurrent Neural Networks
Deep vs. shallow networks : An approximation theory perspective
Decoupled Neural Interfaces using Synthetic Gradients