Research Engineer position - University of Rouen
Main subject: Image Processing - Signal Processing
Secondary subject: Eye Tracking – Simultaneous Localization And Mapping (SLAM)
Place: CETAPS (http://staps.univ-rouen.fr/le-laboratoire-cetaps-191625.kjsp?RH=1379595481948), Faculty of Sport Sciences, University of Rouen, France and LITIS (http://www.litislab.eu), National Institute of Applied Science, INSA of Rouen, France.
Starting: 01/04/2017. Length: 36 months.
Salary: 2264 €/month net of charge (indice 600).
Criteria of eligibility: to have a PhD and/or an engineer certificate.
Key words: SLAM, Eye Tracker, Point of gaze 3D tracking, Bundle adjustment
To apply or for further information please contact:
The XTerM project is a multi-disciplinary project, which is concerned with the analysis of human movement to explore behavioural adaptability as a determinant property of expertise (http://media.wix.com/ugd/8749fe_381f82bc74ae49d3b74b2af42d7e6cbc.pdf). It combines knowledge and methodologies in the human movement sciences, computer sciences and applied mathematics. The primary aim is to investigate the functional role of movement variability in natural contexts where uncertainty and temporal pressure are high, requiring that individuals constantly adapt their behaviours in order to respond to existing dynamical and interactive constraints. Studying the functional role of movement variability involves assessing how adaptive is human behaviour by analysing the balance between movement pattern stability (i.e., persistent behaviour) and flexibility (i.e., variable behaviour) relative to a performance context. Specifically, the project explores how experienced and inexperienced individuals in work and sport, such as firemen and athletes (climbers), adapt their visual-motor behaviours in various performance contexts, i.e., when the environmental properties exhibited more or less uncertainty.
The objective of the present work is to track the 3D coordinates of the point of gaze of a climber and the scan path of his/her visual intake during ascension on an artificial climbing wall.
This work is similar to Simultaneous Localization And Mapping:
3D positions, shape and colour of holds on the climbing wall are known, which gives partial information on the wall geometry as well as visual fixations.
Data collected during ascension are:
No real-time needed, processing can be done offline.
The candidate is requested to have a PhD or an engineer certificate in Image and/or Signal processing. An important knowledge of Matlab and Python is primordial. The work will be done in Rouen and supervised by Ludovic Seifert from the University of Rouen and by Romain Hérault from the INSA of Rouen.
(c) GdR 720 ISIS - CNRS - 2011-2015.