If you’ve been trying to eat healthy or keeping a food journal, you must know how hard it can be, to remember everything you eat or drink throughout the day – especially when the eating and drinking happens around others or while you’re busy doing something else.
Researchers at the Carnegie Mellon University might have a solution for you, in their latest innovation – FitByte, a wearable diet monitor that attaches to your eyeglasses.
How Is ‘FitByte’ Different?
The act of eating, isn’t as simple as we think it is.
It involves several actions like bringing the food or drink close to your mouth, sipping, biting or chewing and swallowing and for a sensor to capture accurately what you eat, it needs to focus on all of these. The monitors designed so far, have had a rather narrow approach and recorded only on one of these aspects which made them unable to gather reliable data in noisy, daily-life environments.
FitByte uses a combination of high-speed accelerometers, a number of gyroscopes, and infrared proximity sensors to detect and track hand-to-mouth gestures, chewing and swallowing. A camera at the front of the glasses detects all food intake, including soft things like ice cream and yogurt, detecting which had so far been a technological challenge.
The superior technology used in FitByte allows it to be highly accurate in many everyday situations like when the user is at a meeting, watching TV, snacking alone, at the gym, or hiking outdoors.
Apart from detecting all stages of eating and drinking, FitByte can even track eating behaviors. It can tell you, for instance, in what situations you tend to eat the most, when you’re binge-eating, whether you eat more when others are around or when alone, etc.
With this invaluable information at your fingertips, you can stay mindful of patterns in your eating behavior and stay close to your health and diet goals.
Bedri, Abdelkareem, et al. “FitByte: Automatic Diet Monitoring in Unconstrained Situations Using Multimodal Sensing on Eyeglasses.” Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems. 2020.