Diet Vision
A health and nutrition monitoring app using computer vision.
DietVision is an innovative system that automatically tracks and predicts the nutritional values of food items using images captured by an armband camera. The project utilizes advanced computer vision and machine learning techniques, starting with the InceptionV3 model and planned upgrades to YOLOV8 for improved accuracy in recognizing food items.
Integrating with a Continuous Glucose Monitor (CGM), it provides real-time glucose monitoring and personalized dietary suggestions. The system’s backend is powered by Flask, with the machine learning model deployed on NVIDIA Triton through AWS SageMaker, ensuring scalable and efficient performance.
Future enhancements include the development of a web and mobile application, along with continuous improvements to model accuracy for better food tracking and health management.
- Award: Secured 3rd place in the International Hackathon–Build Your AI Startup, winning cash prizes and being shortlisted for the NEXT Program.
- GitHub: View on GitHub
Demo Images
