- prepare a survey on human action recogniton based on RGB video and audio data; - investigate and propose new solutions to recognize human activities in the presence of strong occlusion, overcoming the limitations of current systems; - carry out an experimental comparative study of the proposed solutions and state of the art, based on project datasets or publicly available;
Student in an Master in Electrical Engineering, Informatics, Bioengineering or similar
Minimum profile required
- global grade average in study cycle or in the first 3 years of the study cycle greater than or equal to 16 (in 20).
- experience in computer vision; - experience in machine learning models; - experience in deep learning models; - experience in python.
Since 16 Mar 2022 to 29 Mar 2022
Cluster / Centre
Networked Intelligent Systems / Telecommunications and Multimedia