Logging Your Food Is Now Easy as Snapping a Photo

Snap It/photo provided

Tracking your food can be a tedious process. By the time you’ve logged every last ingredient from every last mouthful of your breakfast, it’s practically time for lunch.

Lose It, a Boston-based nutrition app, is introducing a functionality that aims to do away with those little headaches. Snap It allows calorie counters and nutrition neurotics to simply photograph a meal, upload the picture to the app, and let the system populate its nutrition information.

“It’s bringing mindfulness to the way people think about their food and their exercise, and giving them more control by making them smarter and helping them make better choices,” CEO Charles Teague says. Snap It, he says, will make food logging faster, easier, and more effective than it is under existing platforms.

Skeptical? Teague explains that the system uses deep learning to make educated guesses about what you’re eating. The platform uses a “neural network” honed by guessing the content of thousands upon thousands of images, and having those guesses corrected when necessary. Eventually, the technology remembers certain content, and learns to better identify what’s in an image. “The basic idea is you train it with data,” Teague says.

Snap It can’t get as granular as you could when manually entering your food—it can’t tell, for example, how much salt or butter you put in your pasta—but it gives consumers a good idea of what they’re consuming, with minimal effort. (Users also have the power to correct the system if it’s way off base, and to adjust serving sizes and ingredients.)

“It can’t tell every ingredient in your salad, but it can tell that you had a salad,” Teague explains. “[It says], ‘I know it’s a salad, and based upon what other people are eating, and what the most popular salads are, I’ve got some intelligent guesses about how I can match that up with a specific food.'”

Snap It beta is rolling out Thursday as a free—at least for now—part of the Lose It app. The technology is still in its early stages, but Teague says releasing it now will only improve the system.

“From the beta that we’re running, we’re going to get hundreds of thousands of millions of new photos and correlated foods, and we’re going to use that to continue to train and improve the intelligence of the system,” he says. “We think it will just keep getting better and better and better.”