Before you touch a fork to that perfectly styled plate, you know you’ve got to snap a shot and Insta that (d)ish. #FoodPorn.
Google researcher, Kevin P. Murphy — a scientist who specializes in artificial intelligence, algorithms and theory — is developing what he calls Im2Calories. It’s function: to determine the calorie counts of foods in photos on the Internet or social media. It’s purpose: TBD. In actuality, though, it’ll save calorie-counters and Weight Watchers the time of logging individual ingredients.
In Boston last week at the Rework Deep Learning Summit, Murphy revealed his new tool. He explained that Im2Calories will “be capable of counting the number of calories in every item…simply by examining a photo.” Doesn’t even need to be hi-res. Through “sophisticated deep-learning algorithms,” it can identify food items, judge portion size and calculate a calorie count just by analyzing the depth of each pixel. Murphy admitted that the technology isn’t foolproof just yet, and may mistake a poached egg for fried, but it’s getting there.
Now, the FDA’s expansion of calorie counts back in 2014 — placing requirements on all chain restaurants, pizza places, and even movie theaters — was a blessing and a curse. But, it seems sacrilege on my social media posts.
#FoodPorn is supposed to be a guilty pleasure. In regard to calories: don’t ask, don’t tell. Am I right?
Steph is a native New Yorker with a love of culture, creativity, and communications. She’s a foodie, nutrition junkie, yogi, wine-lover, bookworm, world traveler, style-addict, and people person. She also runs her event planning company, Orchard + Broome.
