Abstract:
The objective of this thesis was the implementation of a web application designed for facilitating the deployment of and use them for real time face/recognition.
The key points of this work were:
o Developing a web application that exploits the protocols and standards of WebRTC for the
purpose of transmitting a webcam video and se OpenCV's image processing functionalities
on each frame of the video.
o Containerization of the application with Docker.
o Deploying the application using cloud services.
In the frst chapter, we talked about software development infrastructure and its evolution
and have given a general idea on cloud computing as an infrastructure and the di?erent types of
services it provides. We also talked about machine learning engineering and MLOps. The second
chapter provided an assessment of the system's requirements and described its expected behaviour.
We mentioned all the ressources and tools we have used in the third chapter. Finally, the fourth
chapter illustrated the di?erent interfaces of the system.
In conclusion, the research carried out in this master's thesis addresses the feasibility of developing a system that facilitates the deployment of custom machine learning models for face
recognition, using cloud services as an infrastructure.
