Valais/Wallis AI Workshop

Program

08:30 - 09:00

Welcome Coffee and croissants

09:00 - 09:10

Opening words

09:10 - 10:00

Keynote presentation : Prof. Marcel Salathé

Marcel Salathé is a professor of life science and computer science at EPFL, where he heads the Digital Epidemiology Lab. He is the Academic Director of the EPFL Extension School. He is a co-founder of multiple startups, and advises companies and governments on digital technologies.

Digital cohorts - i.e. cohorts where most data is transferred digitally through sensors, mobile apps, and web applications - are gaining ground. I’ll present our experiences from the Food & You cohort (www.foodandyou.ch), a personalized nutrition cohort in Switzerland, and discuss the potential, and the challenges encountered. The Food & You cohort required us to build substantial digital food infrastructure, including the AI-based food tracking app “MyFoodRepo", and the open food database “Open Food Repo”. I’ll give a short overview over all of these projects, with a focus on the challenges of food recognition using machine learning and crowdsourcing.

10:00 - 10:20

Prof. Dominique Genoud (HES-SO Valais-Wallis, institute of information systems) - Some flying examples of machine learning usage in e-agriculture

10:20 - 10:40

Juan Herrara (Agroscope) - Agronomy from the sky: Opportunities and challenges for integrating image based assessments into an agricultural research program

10:40 - 11:10

Coffee break

11:10 - 11:30

Florian Labhart (IDIAP) - Detecting alcohol consumption through smartphone sensors: examples from the Youth@Night study

11:30 - 11:50

Calixte Mayoraz (HES-SO Valais-Wallis, institute of information systems) - Automated rice grain quality control with machine learning

11:50 - 12:10

Martijn Weterings (HES-SO Valais-Wallis, institute of life technologies) - Intelligent food and food production

12:10 - 12:30

Olivia Mariani (IDIAP) - Can we reconstruct a movie from scanning-aberrated images of the beating heart ?

12:30 - 13:30

Lunch

Registration : https://doodle.com/poll/yxedsbxudhp7d5gx