It is proposed a way to detect wandering activities on alzheimer patients by their movimentation, collected in a non intrusive way.
This resources come from the paper we've published on HEALTHINF 2021 conference
Generated data on the paper can be found here.
The source-code of the generator here
Where the technical part of the deep learning is, the build and train of the convolutional neural network (CNN). Check it out by clicking here
A python package was made to abstract some of the pre processing and model run. For more details check it here
It is planned in the future to have an api server with the model running (Flask). Check it out by clicking here
In order to make it easier to join the features and fixes developed in the project, it was grouped with releases, you can see them by clicking here.