This project was developed during the Summer Innovation Week at CRITICAL Software (2018).
The project focused on the creation of an inteligent lens prototype capable of recognizing humans amongst other visual elements and identifying faces with high precision.
The result is a web service with a live stream of the video processed by the AWS camera and broadcasted to the clients via a server. The viewer can then choose whether to see how many elements are detected by the camera (including simple objects such as chairs and tables, as well as humans) or who's face is looking at the camera (giving the person's name if he/she is in the database).
The base code, the pre-trained deep learning models and the AWS camera were provided by the project's mentors.
/fikalab-deeplens-project - system logic
/fikalab-deeplens-server - web service for data visualization
The original code was written by Jason Bolito, Diana Mendes and Samuel Pedro.
The remaining code and modifications were made by Filipe Pires, Hugo Marques, Inês Valentim, João Martins and João Carvalho.