Google has big design to reinvent the food diary — by get it do most of the work for you . Rather than having to reckon up the Calorie for your hamburger , fries , and heavy soda yourself , their new undertaking , Im2Calories , take to automatically cater large calorie reckoning . An article inPopular Scienceexplains that the system can gauge “ the sizing of each bit of food , in relation to the scale , as well as any condiments " by comparing the pixels in any picture you provide to its pre - existing database of food look-alike and associated calorie tally .
If you ’re worried about accuracy , Im2Calories does countenance users edit the automated results . What ’s more , the organisation is " designed to improve itself " through the cognitive process of “ deep learning . ” If , read Google scientist Kevin Murphy , " it only work 30 percent of the time , it ’s enough that people will start using it , we ’ll collect data point , and it ’ll get proficient over time . "
Public wellness expert and epidemiologists , he tell , are eager to put Im2Calories to work in ordering to cut through vogue across populations , as they look for solutions to our country ’s obesity epidemic .
white potato envisions other software for this form of artificial intelligence as well , let in traffic analysis . By prove street view , “ We do n’t want to just say there are cars in this intersection , " Murphy aver . " That ’s boring . We want to do things like localize cars , number the cars , get attribute of the railroad car , which way are they facing . Then we can do things like traffic vista analysis , forecast where the most potential parking touch is . And since this is all learned from information , the technology is the same , you just change the data . ”
[ h / tPopular Science ]
