A Deep Learning Approach Towards Indian Culinary Classification

Abstract

With the rapid development of deep learning, food image classification has gained significant attention due to its potential applications in dietary monitoring, mobile health, and food tourism. This paper presents an efficient approach for classifying Indian food items using the Mobile Net architecture, known for its speed and performance on resource-constrained devices. The
proposed model was trained on a dataset comprising ten Indian food categories and achieved an impressive
classification accuracy of 98%. The results highlight Mobile Net’s suitability for real-time and mobile-based food recognition systems, especially in diverse culinary environments like India.