IoT Project : Connected Mirror

Presentation :

We developed this project as part of our engineering studies at Polytech Sorbonne, in Embedded Systems specialty.

We had to create a product in the field of IoT that met a need. We came up with the idea of creating a connected mirror, aimed to help people with dermatological diseases like acne or psoriasis. It's a recurrent and very constraining problem for young people who nowadays really take care of their appearance.

The goal was to create a mirror with an easy-to-use interface, to see the advanced of the disease of the client and its evolution using a database, and motivate him to take his treatment if no change is seen by the AI.

The mirror works as follows:

Hardware part : specifications

List of needed objects for this product:

Software part

We implemented a program in Python, with OpenCV, to detect faces.

After this, we save the image to analyse it, and send it to a server with a neural network with DermNet database which is a database of dermatological disease.

The neural network is using PyTorch framework.

We are using QT5 for the user interface, which is very simple and interactive. We can start the program with it, and take a picture by simply clicking on the "send" button.

After analysis, the results are shown on the interface, with advice for the treatment and motivation for the users.

Source AI :

https://github.com/mhw32/derm-ai