Because hiring a nutritionist to judge the health value of everything you eat is hardly a reasonable option, Harvard researchers Jon Noronha and Eric Hysen developed PlateMate, an intuitive application that automatically lets you know how healthy your meal is.
Recent studies show that crowdsourced data is typically as reliable as the professional opinion of a trained weight-watcher, yet gathering information from crowds is more affordable and applicable on larger scales. The process starts off with the user snapping a photograph of their meal, which is then transferred to 'Turkers,' members of an online network established by Amazon to improve product listings. The PlateMate asks 'Turkers' to judge what food is on the plate and in what quantity (a task still impossible for most machines). Once the meal has been analyzed, a program automatically deduces its caloric and health values, transmitting the data back to the original user.
And voila, PlateMate leaves the end user with valuable information, at a fraction of the cost and with all the convenience of Google Maps!