Sign In

Communications of the ACM

ACM TechNews

Crowdsourcing Nutrition in a Snap

View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook
plate of food

Credit: Dr. Jay Heller's Blog

Harvard University researchers have developed PlateMate, software that enable users to utilize crowdsourcing to determine the calorie levels of their meals. Users take a picture of each meal and submit it to the crowd, which estimates the calorie level. In testing, PlateMate's calorie estimates have proven to be just as accurate as those of trained nutritionists, and more accurate than the user's own logs.

"Estimating the nutritional value of a meal is a fairly complex task, from a computational standpoint, but with a structured workflow and some cultural awareness, we've expanded what crowdsourcing can achieve," says Microsoft's Jon Noronha, who co-developed PlateMate as an undergraduate student at Harvard. PlateMate, which works in conjunction with Amazon Mechanical Turk, divides nutrition analysis into several tasks, asking groups of Turkers to distinguish between foods in the photo and estimate the calorie values. The researchers designed simple, clearly defined tasks, and algorithms that compare several answers and choose the best one.

"What makes the nutrition application so interesting as a problem in crowdsourcing is that computers are so very far away from doing it on their own — because food is such a human thing," Noronha says.

From Harvard University
View Full Article


Abstracts Copyright © 2011 Information Inc., Bethesda, Maryland, USA


No entries found