Annonce

Les commentaires sont clos.

PhD position in Computer Vision "Human gestural analysis based on fine-grained motion and graph convolutional networks"

30 Mars 2022


Catégorie : Doctorant


PhD Project short description

The aim of this PhD thesis is to develop systems based on fine-grained motion analysis (using cameras and/or non-intrusive sensors) and graphical convolutional networks (GCN), allowing to accurately detect the gesture of a person in a video stream. The objective of this thesis is to propose a robust and easily adaptable mathematical tool to encode, at different levels of granularity, the movement of an individual, based on the physical constraints induced by the human body (skeleton, muscle). In addition to gesture detection, attention will be paid to motion filtering and temporal segmentation of motion activation so that the system can be easily deployed in a less controlled acquisition context. This work can be applied to different applications, such as home care, medical diagnosis and rehabilitation after an accident, or augmented reality/virtual reality interaction.

Working Environment

The PhD thesis will be hosted in the Centre for Education, Research and Innovation (CERI) Digital Systems at IMT Nord Europe. The CERI SN covers a wide disciplinary field linked to constrained systems (the Internet of Objects, robotics), Humans (and in particular their interactions with the digital world) or even complex systems through the prism of Artificial Intelligence and Automation. The PhD thesis will be conducted within the HIDE group. The aim of this group is to develop models and methods for digital simulation, machine learning or even decision making to help humans reason better, improve their interaction with their environment and better optimise certain processes.

The PhD will be funded for 3 years (1400-1600 € net per month). There will be opportunities to teach at IMT Nord Europe.

 

References

Heidari, N., & Iosifidis, A. (2021, January). Temporal attention-augmented graph convolutional network for efficient skeleton-based human action recognition. In 2020 25th IEEE International Conference on Pattern Recognition (ICPR) (pp. 7907-7914).

Xu, W., Wu, M., Zhu, J., & Zhao, M. (2021). Multi-scale skeleton adaptive weighted GCN for skeleton-based human action recognition in IoT. Applied Soft Computing, 104, 107236.

 

Candidate profile

Qualification: The high-ranked candidate is expected to have a MSc degree, completed by September 2022, with background in image processing, computer vision or machine learning applied for vision.

Experience: The ideal candidate should have some knowledge and experience in Computer vision and in Machine and Deep learning. As for generic competences, we seek a qualified professional, with a teaching vocation, empathy, capacity for teamwork, motivation for innovation, capacity to adapt and to identify with UD’s identity and mission.

Language Skills: Fluent written and verbal communication skills in English are required.

How to apply

Please send by April 15th 2022: CV + grades + copy of degrees + motivation letter to Hazem Wannous, hazem.wannous@imt-nord-europe.fr and Benjamin Allaert, benjamin.allaert@imt-nord-europe.fr