Intelligent appliances with multimedia capability have been emerging into our daily life. Our home is one of the most prominent environments where intelligent appliances and smart wearables are inter-connected to deliver personalized healthy choices. The modern lifestyle isn’t allowing people to spend time on cooking healthy food at home and develop personalized wellness plans. The AI powered personalized wellness system is architected to delivering an immersive experience for an enjoyable and healthy lifestyle.

The system projects an user-friendly screen on any surface or table and turn the surface into a touch screen by using advanced image processing techniques. It assists users to choose the diet preferences based on personalized needs such as low-fat, gluten-free, etc. The system detects any ingredients placed on the table and recommends recipes based on ingredients and diet preferences in real time. The system integrates with smart wearables to provide a comprehensive view of personalized wellness profile including calories, weight, and other parameters for sustainable healthy benefits.

This fully integrated AI edge system is implemented with deep learning at the cloud and applied pruning techniques at the edge to perform rapid model training and real-time inference. A Nvidia Xavier, one of the world’s fastest AI-at-the-edge processors is used for the system deployment to realize GPU accelerated image processing computation and machine learning.

The technical competencies include software tools such as TensorFlow, Python, Flask, OpenCV, NumPy, JavaScript, ReactJS, jQuery, Electron, WebRTC, and AWS SageMaker instance. The hardware stack includes Nvidia Xavier processor, an RGB-Depth camera, and a projector. AWS SageMaker instance is used for cloud deep neural network model training. The instance specification includes 8 vCPU, an Nvidia Tesla V100 GPU, and 61GB memory.