A Machine Learning and Augmented Reality based Framework for Multilingual Product Identification in Retail using Mobilenets and Vuforia

In the modern world, Machine Learning and Augmented Reality have taken the retail industry by storm. Machine Learning and Augmented Reality have provided a major boost to the industry of interactive retail by providing features such as real-time product detection and identification. The proposed research aims at overcoming several challenges in the present scenario which include the time consuming process of standing in long queues while purchasing the products at supermarkets, personalizing the shopping experience in order to maintain the privacy of the users, helping the customers to maintain their specified budgets, reducing the high labor costs and overcoming the language barriers while pertaining to selling products to groups with different linguistic backgrounds by combining the real-world interaction of Augmented Reality and Machine Learning-based product identification. This proposed research work aims at providing the customers with a futuristic shopping experience while maintaining their specified budgets. The Machine Learning-based object detection approach detected the products with around 96% accuracy and the Vuforia-based Augmented Reality approach detected objects with maximum accuracy.