Project

The Hungry Robots Lab

The Hungry Robots Lab at Wageningen University & Research strives to innovate nutrition research. We want to measure what you eat, when you eat and how you eat fully automatically. Our favorite techniques and skills include machine learning, machine vision, deep learning, engineering and magic.

The Hungry Robots Lab is part of the division of Human Nutrition & Health and based on the campus. 

We are part of education, including the new master of data science for health, the minor quantified self, and a number of other courses on data science.

More about Hungry Robots Lab

For more information and latest updates, visit the website Hungry Robot

Overview of research:

smart_tray_333.png

The smart tray

We are using a smart tray with pressure sensors to measure portion and bite-size. The tray looks like a regular dining tray, find more information here at the published article.

machine_vision_333.jpg

Machine vision

We are using facial landmark detection as well as deep learning to real-time analyze chewing behavior and eating rate. 
This can be useful both for regular people and understanding how they eat, as well as for specific situations such as underweight elderly.

food_recognition_333.png

Food recognition

We are using different techniques, including hyperspectral analysis (this definitely borders on magic) and machine learning to recognize the food that is on the plate.