Posts

Accelerating Eulerian Fluid Simulation With Convolutional Networks

/in /by

Jonathan Tompson from Google and his colleagues, Kristofer Schlachter, Pablo Sprechmann and Ken Perlin from the New York University have come up with a really nice technique to teach a convolutional neural network how fluid and smoke simulations work. The project webpage is available here, or you can click on the image below to access the paper.

Status: accepted to ICML 2017

(please note that I am in the acknowledgements section and was not a co-author of the paper)

cnnfluid-cover

Two Minute Papers – Creating Photographs Using Deep Learning

/in /by

Two Minute Papers – Deep Neural Network Learns Van Gogh’s Art

/in /by

Artificial neural networks were inspired by the human brain and simulate how neurons behave when they are shown a sensory input (e.g., images, sounds, etc). They are known to be excellent tools for image recognition, any many other problems beyond that – they also excel at weather predictions, breast cancer cell mitosis detection, brain image segmentation and toxicity prediction among many others. Deep learning means that we use an artificial neural network with multiple layers, making it even more powerful for more difficult tasks.

This time they have been shown to be apt at reproducing the artistic style of many famous painters, such as Vincent Van Gogh and Pablo Picasso among many others. All the user needs to do is provide an input photograph and a target image from which the artistic style will be learned.

Two Minute Papers – Artificial Neural Networks and Deep Learning

/in /by

Artificial neural networks and deep learning provide us incredibly powerful tools in machine learning that are useful for a variety of tasks ranging from image classification to voice translation. So what is all the deep learning rage about? The media seems to be all over the newest neural network research of the DeepMind company that was recently acquired by Google. They used neural networks to create algorithms that are able to play Atari games, learn them like a human would, eventually achieving superhuman performance.

Deep learning means that we use artificial neural networks with multiple layers, making it even more powerful for more difficult tasks. These machine learning techniques proved to be useful for many tasks beyond image recognition: they also excel at weather predictions, breast cancer cell mitosis detection, brain image segmentation and toxicity prediction among many others. If you would like to know more about neural networks and deep learning, make sure to check out this talk from Andrew Ng.

In Two Minute Papers, I attempt to bring the most awesome research discoveries to everyone a couple minutes at a time.

 

Fun with DeepMind’s Deep Q-learning

/in /by

I have had an awful lot of fun with Google DeepMind’s Deep Q-learning algorithm. It plays Atari Breakout solely based on relying the sensory input, and doesn’t know anything about the game when starting out.

I have also added a patch to fix the visualization when reloading a pre-trained network. The window will appear after the first evaluation batch is done (typically a few minutes). This configuration is able to run Ilya Kuzovkin’s version using less than 1GB VRAM.