Recrutement
Ingénieur-PostDoc
POSTDOCTORAL FELLOWSHIP “DATA SCIENCE APPLIED ON POSITIONING ANOMALY DETECTION AND UNCERTAINTY ESTIMATION”
Context
Financed by the French Research Agency (Agence Nationale de la Recherche), this postdoctoral research fellow position will work in the framework of the project ReSilientGAIA (Reliable Positioning System for Soft Mobility Safety Enhancement with a Green AI Approach). The objective of this project is to add safety control in a new reliable positioning algorithm based on multisensory fusion for soft mobility users (pedestrian, bikes, e-scooters, etc). The sensors considered in this project include Global Navigation Satellite Systems (GNSS), Inertial Navigation System (INS), magnetometer and barometer. Our ambition is to design new approaches based on AI to build physical models that describe the signal perturbations in complex urban environments. By improving the safety and reliability of the positioning system for soft mobility, this project contributes to the mobility transition strategy toward lower carbon emissions.
Missions
The postdoctoral fellowship will work on the following missions:
- Vehicular AI model training for GNSS measurement anomaly detection and positioning uncertainty estimation. To do this, the existing labeled dataset will be used in priority. Potential environmental features (such as urban morphological indicators) will be considered using the 3D city model.
- Propose an approach of automatic data augmentation and continuous learning for the pre-trained vehicle anomaly detection model.
- Soft mobility data labeling technique development using fisheye camera: propose a method to further evaluate and improve the accuracy of the labeling technique developed in [1], which uses image segmentation and satellite projection.
- Once the labeling technique is developed, a soft mobility multiple-sensor dataset will be constructed. The soft mobility considered in the project includes mainly pedestrians, bikes and e-scooters.
- Analyze the feasibility and the possible approaches of the transfer learning between vehicle and pedestrian.
[1] Zhu, N., Bouronopoulos, A., Leduc, T., Servières, M., and Renaudin, V., 2023, April. Evaluation of the Human Body Mask Effects on GNSS Wearable Devices for Outdoor Pedestrian Navigation Using Fisheye Sky Views, In Proceedings of the 2023 IEEE/ION Position Location and Navigation Symposium (PLANS), 24-27 April, 2023, Monterey, United States.
Required Skills
The research work will be co-supervised by the researchers from GEOLOC laboratory at the University Gustave Eiffel and the AAU-CRENAU laboratory of CNRS. The candidate should hold a PhD degree in data science, applied mathematics or computer science. Specialization in Artificial Intelligence (or statistical learning methods) and experience in their application to one or more of the following fields are required: Signal processing, Computer vision, Geomatics, and Navigation.Engineering degree, Master degree or PhD degree in: data science, applied mathematics or computer science or any other comparable discipline.
- Data Science.
- Artificial Intelligence (time series analysis).
- Signal / image processing.
- Knowledge of Python and Matlab coding.
- Scientific writing.
- Enthusiasm, responsibility and excellent collaboration skills.
- Organization and rigor.
- Strong oral and written skills in English.
Contract Type | 24 months full time (38h30/week) with a starting date from February / March 2024 |
Location | GEOLOC Laboratory, University Gustave Eiffel, Nantes, France |
Application | Attach all your materials (Your motivated application, CV, Diplomas, List of publications and productions) in one PDF file emailed to ni.zhu@univ-eiffel.fr |