Food impacts our health. Each of us responds to food in different ways, giving rise to the promise of personalized nutrition. But research is lacking easy-to-use tools to measure what individuals are eating.
To solve this problem, AI For Nutrition supports the development and use of open, AI-assisted food tracking tools for cutting-edge research on food and health. These tools have been developed to minimise the burden on cohort participants and maximise the ease of deep data extraction for researchers.
MyFoodRepo is a mobile app that tracks your food consumption with photos and barcodes using artifical intelligence.
MyFoodRepo has been rolled out in multiple clinical and digital cohorts – such as Food & You – in Switzerland and will soon be available for cohorts in other geographies.
The Open Food Repo is a community driven open database for barcoded food products.
The Open Food Repo currently contains more than 40,000 Swiss products, and an expansion to other geographies is underway. In addition to traditional nutritional values, the Food Repo will also add genetic data to food products in the future.
Image-based food recognition is a challenging AI problem – algorithms need to recognize which parts of an image represent food, what food it is, and how much. The MyFoodRepo AI Benchmark is an open, continuous machine learning competition that uses crowdsourcing to collaboratively develop the best AI for food recognition.
Are you a scientist interested in using our technologies in clinical cohorts, or a foundation eager to be part of the revolution in nutritional epidemiology?