University of Buffalo
The days of manual calorie counting may soon be over, thanks to AutoDietary, a wearable technology that recognizes different types of food by monitoring eating sounds. It provides an automated way of measuring daily caloric intake, according to Food Navigator.
AutoDietary is made up of two parts, an embedded hardware system and a smartphone app. An acoustic sensor, which is worn around the neck, detects eating sounds and transmits data to the smartphone app, which is able to take the data and interpret it as different food types.
Computer scientist Wenyao Xu and team at the University of Buffalo tested the system’s effectiveness by gathering 12 subjects and asking them to eat seven different types of food: Apples, carrots, cookies, potato chips, walnuts, peanuts, and water. The researchers calculated an overall accuracy of 86.6 percent for identifying chewing or swallowing events, and 81.5 percent to 90.1 percent accuracy in recognizing different food types.
Future research will build upon the food library, which will be done by testing different foods and recording the eating sounds made, reports Science Daily. Xu is also planning a biomonitoring device to complement AutoDietary, which would be activated when the user is eating a general category of food.
The device would determine nutritional value based on blood sugar levels and other measurements, rather than relying on sound alone for a more accurate analysis.