Deep Learning-Based Diabetes Detection and Real-Time Exercise Monitoring for Enhanced Healthcare

Diabetes is a major global health issue that requires early detection and appropriate treatment.Since it can't be treated we can make measure and address this issue by merging real-time fitness tracking and deep learning for diabetes detection. We create a robust diabetes detection system with high accuracy by carefully identifying features, using machine learning, and preprocessing data using the explainable boosting model. In parallel, we use Mediapipe to develop pose detection models for real-time evaluation of workout forms, allowing users to receive fast feedback. Exercise selection is also done keeping the injuries patient may have in mind. By integrating these elements, we hope to encourage a healthy lifestyle, empower people to control their health care, and advance early diabetes detection.