Health & Wellbeing

Food-tracking necklace listens to you eat

Food-tracking necklace listens...
A prototype of the Autodietary food monitoring necklace
A prototype of the Autodietary food monitoring necklace
View 1 Image
A prototype of the Autodietary food monitoring necklace
A prototype of the Autodietary food monitoring necklace

The sound that you hear while chewing could become the next way of monitoring your caloric intake. That's the idea behind what researchers from the University at Buffalo in the US and Northeastern University in China are doing – they're creating a necklace-like device that monitors the sound of chewing, and matches that sound to the calories of the food being consumed.

The device is called Autodietary, and the chewing-sounds library that it uses is being developed based on the premise that everything has a unique sound when chewed. The growing library of chewing sounds serves as Autodietary's "brains."

When you wear the Autodietary device and start chewing, a tiny high-fidelity microphone in the necklace records the sound, sends that data to a smartphone app via Bluetooth, where it then compares it to the information in the library. That information is used to confirm what you're eating and determine how many calories you're consuming.

Autodietary has been tested to some effect with the ability to determine the difference between apples, carrots, potato chips, cookies, peanuts and walnuts with about 85 percent accuracy. But the average person's diet is obviously more complex.

The device admittedly has a long way to go before it can determine the difference between the sounds of the same sweetened or unsweetened cereals or whether that burger you're gnawing on is made of lean or regular ground beef. Over time, the researchers hope to refine the necklace and library to the point where it delivers a greater degree of accuracy, perhaps with the addition of some type of biomonitoring device.

Researchers say that the immediate feedback provided by the Autodietary device and accompanying library will ultimately give people suffering from a host of dietary issues like diabetes, obesity, and other ailments the ability to better monitor their food intake and manage their situation.

Source: University at Buffalo

Despite the billions of dollars and thousands of man-years devoted to speech recognition by the major tech firms, the unreliability of such apps still results in embarrassing text messages to random recepients.
In this case, you mean to tell me that this magical algorithm can distinguish the difference between the chewing a piece of chicken that was grilled versus one that was fried in butter? dark chocolate vs white chocolate? Or the drinking of 2 calorie black coffee versus a 2000 calorie mocha-frappa-whippo-pumpkin-eggnog-holiday-diabetes-in-a-cup?
Also, I'm assuming one would have to calibrate the device to each individual by having to eat everything you would possibly eat. How many calories would the calibration process entail?
Douglas Bennett Rogers
It work much better if it flagged high glycemic and inflammatory compounds.