You are viewing content from a past/completed QCon

Presentation: Workshop: Audio Classification with Convolutional Neural Networks


Location: Henry Moore, 4th flr.

Duration: 1:40pm - 3:45pm

Day of week: Wednesday

Share this on:


The world is full of sound, and until recently, it was a solely human characteristic to be able to identify from whom or to what those sounds belong. But not anymore! With the development and increased availability of machine learning models, accurately identifying sounds is an ability within the grasp of modern computers, and programmers.

In this workshop, we will cover the basics of convolutional neural networks, and then we will train a convolutional neural network (powered by IBM Watson) to identify spectrograms of a number of speakers - namely - late night hosts Conan O’Brien and Stephen Colbert.  Please bring your laptop for the hands-on lab.

Speaker: Sean Tracey

Developer Advocate @IBM

Sean is a technologist and published author. When he's not making something with computers, he usually writing about them instead.

Find Sean Tracey at

Last Year's Tracks