Open Cotutelle PhD position in the field of data science for remote sensing applications between the University of Rennes 1 (IETR UMR CNRS 6164 - MULTIP, Lannion, France) and the Norwegian University of Science and Technology (NTNU, Colourlab, Gjøvik, Norway) on dimensionality reduction and semantic classification techniques to detect and monitor harmful algal blooms in Breton and Norwegian waters based on multi-scale, graph-based, and unsupervised learning methods.
A PhD position in the field of data science for remote sensing applications is available as a co-tutelle between the University of Rennes 1 (France) and the Norwegian University of Science and Technology (NTNU). The successful candidate will be awarded a PhD degree from both Universities. The position is primarily based at the IETR laboratory (MULTIP research team) in Lannion, but the selected candidate will also spend significant time at the Colourlab in Gjøvik, Norway. The candidate will collaborate with a team of experienced researchers from both the IETR and the Colourlab.
Hyperspectral imaging (HS) is a widely used technology in various environmental applications. It provides access to physical and biological properties of imaged scenes, enabling for example the detection and monitoring of algal blooms in lakes and coastal waters. HS has become more prevalent due to the development of increasingly affordable and effective sensors, as well as the utilisation of drones in remote sensing applications. However, the high dimensionality and complexity of HS data present challenges for processing and analysis. Supervised methods necessitate a substantial volume of annotated data for calibration or training, but obtaining accurate ground truth data is difficult and costly. On the other hand, unsupervised methods can identify useful information without relying on ground truth data, making them cost-effective and practical.
This thesis project will focus on dimensionality reduction and semantic classification techniques to detect and monitor harmful algal blooms in Breton and Norwegian waters based on multi-scale, graph-based, and unsupervised learning methods.
The ideal candidate will have a solid background in signal processing applied to remote sensing. She/he should have a strong interest in statistical analysis and data science/machine learning along with proficient programming skills.
Selection will be based on motivation, adequacy of profile and experience, as well as grades. Furthermore, the selected candidate must pass security clearance procedures at both universities.
The application must include:
• A cover letter where the applicant describes his/her personal motivation and relevance with respect to the requirements of the position.
• Transcripts and diplomas for bachelor's and master's degrees (or an official letter stating the approximate date of graduation).
• Name and contact information of three referees
with the subject “Application for PhD in Unsupervised Learning for Hyperspectral Remote Sensing”.
Note that applications will be reviewed on an ongoing basis so you are encouraged to apply early, and no later than November 31st.
Supervisors : Kacem Chehdi (IETR, Multip, UR, France) / Steven Le Moan (Colourlab, NTNU)
Co-supervisors: Benoit Vozel et Claude Cariou (IETR-Multip, UR)