Merging features to classify different levels of Alzheimer's disease dementia (MSc thesis)

As part of my final year project at Ecole Nationale Polytechnique, we carried out a project aimed at detecting and classifying different levels of Alzheimer's dementia based on brain MRI images. We implemented an innovative architecture based on feature fusion between a HOG and a CNN, followed by a KNN classifier, achieving results that exceed the state of the art.

Smart Hive

In this project, we presented a prototype of a connected beehive based on the LoRa and MQTT protocols. We developed an IoT system based on several sensors. Namely, temperature, humidity and light, with the aim of placing them inside the hive to establish a complete monitoring system. To do this, we used the Arduino environment for the sensor part and MQTT transmission to a Raspberry Pi, which acts as the central computer and will also manage LoRa transmission to a third-party platform to display a monitoring dashboard for the beekeeper. This project is preliminary work, and a physical implementation is envisaged for field testing within a set of connected beehives.