Predicting Alcoholic Percentage of a Cocktail Using Deep Learning & Image Processing

Abstract

Alcohol consumption has been practiced as a social norm of relaxation and method of connecting to others by most people since generations. However, there are plenty of health risks involved with such habits if each individual isn't aware of the amount of alcohol they consume and unable to control their intake. In the meantime, food photography has become quite a popular activity especially among young users on the internet. Image recognition is a concept that has been developed in AI models for sometime now and that concept can be used to connect the problem to a novelty solution. This study focuses on developing an application which facilitates the user to input a photograph of a cocktail which will be processed by the system to generate a report that gives out important information like alcoholic percentage of the drink and cocktail recipes.