It’s the dream of any foodie who watches what they eat to be able to snap a photo of their meals and have their phone instantly tell them how many calories they’re about to consume.
That’s the mission statement of a new startup called AVA, which promises to do away with the dreary manual logging process of rival healthy-eating apps in favor of an altogether more streamlined process. Using AVA’s “intelligent eating” service, users will simply take a photo of their food, text it to AVA, and then receive health and caloric information in return.
“We’re using artificial intelligence to assist nutritionists in estimating calories as well as making recommendations, factoring in historical eating habits, diet patterns, location and behavioral analysis against a database of roughly 50,000 meals,” Ian Brady, AVA’s co-founder and CEO, tells Digital Trends. “We’ve seen in our early pilot that factoring in larger data sets makes for more accurate and personalized recommendations, and that these play a considerable role in driving engagement and overall effectiveness of our programs.”
Since AVA is still in private beta mode, Brady’s not spilling the beans on exactly how the technology works but he says that it’s a “combination of image recognition, human recognition and AI algorithms.”
In addition to simply recognizing what users eat, AVA is also hoping to go one step further by making personalized recommendations for future meals — along with ongoing support from (presumably human) AVA coaches.
It’s definitely an intriguing idea, and one that could be absolutely massive if executed correctly. Whether it’s Apple’s WWDC announcement about using artificial intelligence to categorize your holiday snaps or Facebook learning to recognize our faces, AI has made big leaps forward in image recognition in recent years. Already AVA has impressed the likes of former Google CEO Eric Schmidt, whose Innovation Endeavors is one of several venture firms to have invested a total of $3 million in the project.
We’re hungry to hear more… calorie dependent, of course.
