M/F: PhD student : Statistical learning for the observation and modelling of displacements using aerial images
6 Apr 2024
Job Information
- Organisation/Company
CNRS- Department
Laboratoire d'analyse et d'architecture des systèmes- Research Field
Engineering
Computer science
Mathematics- Researcher Profile
First Stage Researcher (R1)- Country
France- Application Deadline
29 Apr 2024 - 23:59 (UTC)- Type of Contract
Temporary- Job Status
Full-time- Hours Per Week
35- Offer Starting Date
1 Oct 2024- Is the job funded through the EU Research Framework Programme?
Not funded by an EU programme- Is the Job related to staff position within a Research Infrastructure?
No
Offer Description
The thesis is jointly funded by Cerema, the MIDOC federation (Intelligent and Sustainable Mobility in Occitania) and LAAS-CNRS.
o LAAS-CNRS is a CNRS unit attached to the CNRS Computer Science and CNRS Engineering institutes. Located in Toulouse, it is associated with six founding members of the University of Toulouse.
o Cerema is a public administrative body (2,600 employees) based in Lyon-Bron, under the supervision of the ministries responsible for sustainable development, transport, town planning and housing.
o The MIDOC federation is one of 15 key challenges funded by the Occitanie region. It is made up of 20 laboratories or research centres in Occitanie.
LAAS-CNRS (employer) and Cerema are combining their skills to supervise this work. The person will be based at LAAS-CNRS and Cerema, a direct neighbour on the Rangueil University Campus in Toulouse.
The position is located in a sector covered by the protection of scientific and technical potential (PPST), and therefore requires, in accordance with the regulations, that your arrival be authorised by the competent MESR authority.
In view of global warming, we urgently need to adapt our cities to the safe development of active modes of transport, in particular by adapting our infrastructure. To do this, we need to be able to provide decision-makers with planning recommendations tailored to each situation. We therefore need to equip ourselves with the means to observe and collect data, so that we can analyse, understand and model journeys.
Modern tools, such as drones and tethered balloons, enable aerial photography which can provide information about the use of a road infrastructure. Their implementation requires the development of algorithms to detect and classify moving objects, which, combined with movement prediction methods, enable a quantitative analysis of the risks associated with uses. The data will consist of images recorded by cameras on board a tethered aerial balloon developed by Cerema (see background).
The proposed work is structured along two lines:
o The first area of research involves building high-performance learning algorithms specifically dedicated to the problem of detecting users using aerial views. The innovative approach based on an aerial balloon will compensate for the weaknesses of the camera approach from the ground or a pole (difficult depth estimation, short range), and will provide better detection performance than existing algorithms using images from surveillance cameras or laser sensors. The results of this first area will provide exhaustive data on the use of an infrastructure (crossroads, roundabout, square, etc.) of a quality comparable to or better than the state of the art.
o The second line of research, based on the previous data or using public data, will consist in designing high-performance algorithms to detect potentially dangerous situations from the road safety viewpoint. Drawing inspiration from recent work in the context of the evaluation of autonomous vehicles, we propose to rely on the combination of probabilistic inference and machine learning techniques. This will involve learning the specificities of the movements of heterogeneous users of an infrastructure, so as to enable a reliable prediction of collisions, and thus better detect dangerous interactions.
The aim of this thesis is to propose new methodological approaches for assessing the performance and risk of road infrastructure, by taking vulnerable users into account. In addition to the theoretical work that will make it possible to achieve this objective, the aim is to produce a new tool for observing urban mobility for use by researchers, managers and decision-makers. To our knowledge, there is no equivalent approach.
Requirements
- Research Field
- Engineering
- Education Level
- PhD or equivalent
- Research Field
- Computer science
- Education Level
- PhD or equivalent
- Research Field
- Mathematics
- Education Level
- PhD or equivalent
- Languages
- FRENCH
- Level
- Basic
- Research Field
- Engineering
- Years of Research Experience
- None
- Research Field
- Computer science
- Years of Research Experience
- None
- Research Field
- Mathematics
- Years of Research Experience
- None
Additional Information
Additional comments
Profile required: Graduate (or able to graduate in 2024) of a Research Master's degree (or equivalent) in Applied Mathematics or Statistics or Business Intelligence, or Artificial Learning and Intelligence, or Signal Processing.
o Skills in software development and databases.
o Good general scientific background, particularly in computer science, machine learning (AI) and stochastic models.
o Good communication skills in French and English (oral/written).
o Qualities required: highly motivated, autonomous, rigorous, able to put forward proposals, open to multidisciplinary approaches.
Applications must include :
o A curriculum vitae and, if possible, at least one letter of recommendation or reference
o A covering letter describing the candidate's interest in the position
o Master's grades (1&2)
o An example of written work (thesis, article, report, etc.)
- Website for additional job details
https://emploi.cnrs.fr/Offres/Doctorant/UPR8001-PATDAN-005/Default.aspx
Work Location(s)
- Number of offers available
- 1
- Company/Institute
- Laboratoire d'analyse et d'architecture des systèmes
- Country
- France
- City
- TOULOUSE
- Geofield
Where to apply
- Website
https://emploi.cnrs.fr/Candidat/Offre/UPR8001-PATDAN-005/Candidater.aspx
Contact
- City
TOULOUSE- Website
http://www.laas.fr
STATUS: EXPIRED