This post is another collaboration with Jan Schlüter from the OFAI (@f0k on GitHub), a fellow MIR researcher and one of the lead developers of Lasagne. He recently added a cool new feature that we wanted to highlight: enabling the use of arbitrary Theano expressions as layer parameters.
UPDATE (April 27th): the paper is now available on the journal website: http://mnras.oxfordjournals.org/content/450/2/1441
The National Data Science Bowl, a data science competition where the goal was to classify images of plankton, has just ended. I participated with six other members of my research lab, the Reservoir lab of prof. Joni Dambre at Ghent University in Belgium. Our team finished 1st! In this post, we’ll explain our approach.
Guest post: Jan Schlüter from the OFAI, a fellow MIR researcher I have met at several conferences, recently added a feature to Theano that fits so well with my previous two posts on fast convolutions that we decided to include his writeup on my blog. So enjoy the third part of the series, written by Jan!
This summer, I’m interning at Spotify in New York City, where I’m working on content-based music recommendation using convolutional neural networks. In this post, I’ll explain my approach and show some preliminary results.
Last month I wrote about how you can use the cuda-convnet wrappers in pylearn2 to get up to 3x faster GPU convolutions in Theano. Since then I’ve been working on an FFT-based convolution implementation for Theano. Preliminary tests indicate that this approach is again 2-4x faster than the cuda-convnet wrappers.
Convolutional neural networks (convnets) are all the rage right now. Training a convnet on any reasonably sized dataset is very computationally intensive, so GPU acceleration is indispensible. In this post I’ll show how you can use the blazing fast convolution implementation from Alex Krizhevsky’s cuda-convnet in Theano.