Tracing Adulteration in Cumin, Cinnamon, and Coffee using Transfer Learning

Food adulteration has been a matter of distress since the dawn of humanity. It not only reduces the quality of food items but also has a variety of adverse health consequences. For value evaluation and consumer safety over deceptive operations, reliable food testing and adulterant screening of various food items are necessary. We establish a system of convolutional neural networks employing transfer learning to differentiate between a food product and added adulterants. Our findings revealed that the newly designed model could accurately and rapidly assess various spices, nuts, and beverages.