Onboard pedestrian tracking for driving assistance

This project aims at detecting potentially dangerous driving situations and initiate countermeasures.

The objectives of the project “Onboard pedestrian tracking for driving assistance” are to increase security through pedestrian detection and trajectory prediction. Based on this prediction, it will indeed be possible to estimate the level of danger that each pedestrian represents, and to infer appropriate actions of the car in due time to avoid an accident. This PhD thesis will develop onboard pedestrian detection, tracking and trajectory prediction, and will be closely related to vehicle command strategies.

Pedestrian detection and tracking is an important research topic in computer vision. Detection and tracking from moving cameras are recognized as extremely complex problems. The Signal Processing Laboratory 5 (LTS5) has a strong and renowned scientific expertise in image analysis, signal and data processing and microelectronics to propose a robust and accurate solution. The Transport and Mobility Laboratory (TRANSP-OR) will bring its unique expertise in terms of pedestrian behavioral modeling.

This project will last 48 months and will be carried out in the context of the MINES ParisTech Chair on automated driving “Drive for you”, sponsored by Groupe PSA, Safran and Valeo. It is jointly led by the Signal Processing Laboratory 5 of Prof Jean-Philippe Thiran and the Transportation and Mobility Laboratory of Prof. Michel Bierlaire.

Principal investigator Prof. Jean-Philippe Thiran
Project manager
Damien Matti
Sponsor MINES ParisTech Foundation (Safran, Groupe PSA, Valeo)
Period 2015-2018
Laboratory LTS5, TRANSP-OR
Industrial partner L.I.F.E. SA
Collaboration TRACE