“The Classyfier” is a table that chooses music to fit the situation happening around it based off of the beverages that the people at the table consume. It chooses a playlist by comparing the characteristic sounds to a catalogue of pre-trained examples. The three classes that the table can detect are hot beverages, wine, and beer. I thought this project was pretty interesting because it is sort of an intro to smart objects and machine learning.
This project was created by Benedict Hubener, Stephanie Lee and Kelvyn Marte at the CIID alongside Andreas Refsgaard and Gene Kogan. They used the OFX collection, Wekinator and Processing to bring the project together.