There are few things in the world worse than trying to order a drink in a busy bar: pushing through the drunken crowd, trying to make eye contact with the bartender, and then shouting a drink order over loud music. You know the dance. And the good news is that Visual computing and A.I. company Nvidia and Herta Security know it well, too, and have teamed up to create a solution: YouBar.
YouBar is a facial recognition software that uses cameras behind the bar to make a record of your face and drink order. The next time that you approach the bar, cameras pick your face out of the crowd and alerts the bartender of your name and last drink order. Viola, your drink is ordered and being prepared by a system with a much better memory than that mustachioed barkeep wearing the suspenders – and before you even reach the moist counter.
Don’t expect to see YouBar at any of the hip dives around the country anytime soon, though. It’s only a concept design (albeit a very popular concept design) at the 2016 Graphics Processing Unit Technology Conference (GTC). Still, considering Chinese robot waiters have already moved from being hired to fired, it’s not hard to imagine YouBar’s tech in the bars of the not-too-distant future.
Herta Security’s video cameras blanket banks, government buildings, train stations, and other population hubs. The company is focused on capturing criminals after they commit crimes, but as every CCTV conspiracy theorist knows, the same video can easily be used for other purposes. Herta CEO Javier Saeta told the Nvidia blog that YouBar is just a fun way for people to get acquainted with facial recognition technology.
Nvidia enters this harmless iteration of surveillance-state actions with its new Jetson TX1 visual computing platform. Jetson TX1 helps Herta’s cameras recognize a person’s face, gender, ethnicity, age, and facial expression. It can also do something called “frontalization,” which moves a person in the video’s face from an angle to straight ahead.
This is the type of computing that could help Herta’s security business, but it could also help them transition into the retail data business. A backlog of big data on each customer’s preferences would be useful for retail and service businesses in more ways than just drink orders, despite how creepy it may sound.
But Orwellian Big Brother fears aside, can’t we all agree the drink-ordering process could use some added efficiency?