In the framework of the research project “Innovative systems and services for transport and production” I-SITE CAP 20-25 (Challenge 2) and the LabEx IMobS3, a postdoctoral position is proposed for highly motivated candidates interested in machine learning. The candidate will be joining the Département Laboratoire de Clermont-Ferrand of the Cerema that benefits from a unique fog and rain platform.
Probabilistic assessment of a detection system under adverse weather conditions
The project "Autonomous VEhicle uNder difficult CondiTions in partially Unknown Road Environment (AVENTURE)" funded by the Challenge 2 of CAP 20-25 aims to develop navigation tools for an autonomous agricultural or road vehicle, making it possible to move, as a human being can, in complete safety from a point A to a point B without first learning the path followed.
An original methodology is developed in Johann Laconte's thesis (2018-2021) on the optimal and safe traversability for the movement of autonomous vehicles. It is based on a dynamic risk map established during the evolution of the vehicle (taking into account obstacles, etc.), based in particular on perceptive sensors.
Adverse weather conditions impact the information delivered by the sensors, and quantifying the impact of this degradation on the risk map is an open problem. The postdoctoral applicant will contribute to this task, focusing mainly on disturbances caused by fog, rain and possibly snow. The first step will be to identify the sensors on which the work will focus (visible, infrared, radar, lidars, etc.), as well as the high-level perception tasks that can be degraded (obstacle detection, pedestrian detection, marking detection,..). Several tests will be conducted within the Cerema's PAVIN Fog and Rain platform in order to adjust probabilistic models on the performance of detection algorithms under adverse weather conditions. The proposed methodology will be compatible with any type of sensor and/or algorithm. This work will be based on those developed in two recent theses co-supervised by Cerema and Institut Pascal, which focus on the impact of rain on image descriptors and weather classification by image processing.
Please send your CV, references and publications by email to:
Research Engineer at CEREMA Centre-Est/Département Laboratoire de Clermont-Ferrand
Phone 04 73 42 10 87
Email : Frederic.Bernardin@cerema.frContinue reading
|Title||Post-doctoral position - Probabilistic assessment of a detection system under adverse weather conditions|
|Job location||IMobS3 Laboratoire d'Excellence - Université Clermont Auvergne 4 Avenue Blaise Pascal TSA 60026 63178 AUBIERE cedex, 63178 AUBIERE|
|Published||June 24, 2019|
|Job types||Postdoc  |
|Fields||Remote Sensing,   Applied Mathematics,   Probability Theory,   Computational Mathematics,   Machine Learning,   Image Processing,   Stochastics  |