A Food Recommendation System for Predictive Diabetic Patients using ANN and CNN
N.C. Brintha, P. Nagaraj, Arige Tejasri, Bhavanam Vijaya Durga, Mederametla Tarun Teja, Maguluri Navi Venkata Pavan Kumar
Abstract
In the present society, many people are suffering from diabetes. To maintain a healthier diet, people must track the number of food calories they consume daily, along with their total calories. There is a growing desire for automated technologies that can help diabetics manage their diet and benefit the population by better analyzing the impact of different types of food on diabetes glucose levels. This research study includes two types of datasets: one for diabetic prediction and one for food recognition. This paper has implemented a diabetic prediction using ANN (3 dense layers), food recognition along with nutrient values by using the CNN model and transfer learning based on VGG16, and a food tracking system to track daily food along with total calories. Finally, our results will be available on the website, which can recognize food and track food items along with estimating calories. With the help of user input based on their glucose readings, the proposed model will predict whether the user has diabetes or not. Based on the user’s preference, they can use a food recognition and food tracking system.