DEEPWine
Virtual oenologist based on deep learning and optical spectroscopy
Auf einen Blick
- Projektleiter/in : Prof. Dr. Francesca Venturini
- Projektteam : Bruno Friedrich Joachim Bender, Ziga Nosan
- Projektvolumen : CHF 49'816
- Projektstatus : laufend
- Drittmittelgeber : Stiftung
- Kontaktperson : Francesca Venturini
Beschreibung
The goal of this project is to perform a feasibility study on realizing of a virtual oenologist, an integrated computational framework to assess wine quality and taste through purely optical analyses. The goal will be achieved by combining two optical spectroscopy techniques (fluorescence and Raman spectroscopy) to gain detailed information on wine components in a non-destructive way, and multimodal machine learning to link the chemical composition to wine quality and taste attributes. This feasibility study will allow to explore the feasibility of an affordable and efficient alternative to chemical analysis through real-time monitoring, which could help wine production practices towards sustainability goals by reducing costs and improving quality control practices.