Applying AI and Big Data to Determine Authenticity of Spices

Abstract
Adding adulterants to spices is potentially a form of food fraud. This paper reports on research which investigates the viability of applying AI methods tothe vast amounts of data that can be gathered using infra-red reflectance spectroscopy, a non-destructive technique, from samples of spices. The goal is to develop a robust practical tool that can be applied to classify samples as having adulterants or not. The AI models can be offered for inferencing on the Cloud or may be deployed on hand held spectrometers. Keywords - Food Authenticity, Adulteration, Spectroscopy, Big-Data, AI Classification, High Performance Computing