Your future #instafood pics could tell you just how many calories are on your perfectly-filtered plate of french fries. At least, if Google has anything to say about it.
Google confirmed that its artificial intelligence researchers are working on an experimental project that can detect your meal’s calorie information by analyzing a photo of the food. The project, called Im2Calories, was introduced last week at Boston’s Rework Deep Learning Summit, according to a report in Popular Science.
The experiment “marries visual analysis — in this case, determining the depth of each pixel in an image — with pattern recognition,” the Popular Science report states. But even though Im2Calories is analyzing pixels, the tech can apparently work with lower quality photos, like Instagram posts, Google research scientist Kevin Murphy said during the event.
The algorithms will become more accurate over time, Murphy said, and could benefit both consumers who are worried about portion control as well as doctors and researchers who could potentially use the data in aggregate for research.
“To me, it’s obvious that people really want this and this is really useful,” Murphy said, according to the report. “OK, fine. Maybe we get the calories off by 20%. It doesn’t matter. We’re going to average over a week or a month or a year. And now we can start to potentially join information from multiple people and start to do population level statistics.”
Google is hardly the first company to work on automated calorie-counting. SmartPlate, an app-enabled plate equipped with scales and sensors that also claims to help measure food intake, debuted on Kickstarter last month. Similarly, San Francisco startup Vessyl successfully crowdfunded a “smart cup” that can track the calories you drink. But Google’s approach stands out as it doesn’t rely on extra hardware to complete its analysis, which could make automated calorie tracking much more accessible and easier to use.
Though Im2Calories is still an experiment, Murphy said the company had filed patents on the technology, so it seems possible the tech could make its way into one or more of Google’s products eventually, once the algorithms have been fine tuned more.