Time: 14:00 - 15:00
Audio fingerprinting: How to make your own Shazam from scratch
Ever heard of "Haar wavelets" or "locality-sensitive" hashing?
Sebastien will explain some techniques used to identify & index audio clips in order to perform live pattern matching. He will demonstrate the identification of audio played from one device's speaker via another device's microphone using a complete dependency free C implementation.
Presented by: Sebastien Paumier, Senior Software Engineer, Mjoll
Venue: Media Lab, 9th floor, tower 3, Media City Bergen
The presentation will also be streamed on IBM: https://video.ibm.com/channel/JwKVCZnuuf5