Thanks to the Internet of Things, breakthroughs in machine learning, and the plethora of cheap sensors that are readily available, virtually every “dumb” piece of household furniture now has a smart counterpart. The latest example of this? A table created by researchers at Denmark’s Copenhagen Institute of Interaction Design that is able to detect what beverages people are drinking around it and select the appropriate mood music.
“Have you been in situations where the mood dropped because someone took forever to decide which music to play?” Kelvyn Marte, one of the three designers (along with Benedict Hubener and Stephanie Lee) who created the table, told Digital Trends. “The Classyfier is a table that automatically chooses music to fits the current situation. It does that by picking up the characteristic sounds of beverages consumed around it, [such as] the clinking of wine glasses or the sound of opening a can of beer. It basically takes the role of the designated DJ so that people can focus on other things. We found it exciting to use sound as an input because it opened up a lot of new interactions that could be either very intentional or completely seamless to your normal behavior.”
In terms of hardware, the table boasts a built-in microphone which picks up sounds. These sounds are then compared to a catalog of pre-trained examples. Depending on whether it detects hot beverages, wine, or beer, it can then select the perfect playlist for each occasion.
Sadly, if you’re thinking this would be the perfect centerpiece for your next drinks soirée, Marte, Hubener, and Lee have some bad news for you: There are no plans to commercialize the Classyfier table right now. “We were [really] thinking about ways to use machine learning in a fun and light way to enhance ambiance in daily situations,” Marte said. “In the beginning, the thought of picking up the sound of opening a beer can was more of a joke, but we kept coming back to it. For now, we are all a bit caught up in other projects, but who knows [if we’ll revisit it later on!] Machine learning is a great tool for designers, and we are excited to find more ways to interact with smart objects using sounds as an input.”
We will drink to that! With the appropriate soundtrack, of course.
