Job Opening

Deep learning methods for automatic mapping of the West African littoral from satellite and topograp

GEOAZUR & LEGOS & LUXCARTA

Posted: 06/17/2019

Open Until: 08/31/2019

Research areas: deep learning, computer vision, geosciences

Position funded by the Institut de Recherche pour le Développement, IRD (https://www.ird.fr/). IRD is a French public research institution that defends an original model of equitable scientific partnership with the countries of the South and an interdisciplinary and citizen science, committed to the achievement of the Sustainable Development Goals.

Project abstract
The West African coast is one of the worldwide littoral areas most threatened by the ongoing climatic and anthropogenic global changes: with a low topography, a weak geological substratum, a poor fresh water supply, and a dense and rapidly expanding population, the West African littoral is highly vulnerable to current sea level rise, extreme climatic phenomena, erosion, and modifications of the ecosystems and resources. The project focuses on the ~3000 km-long littoral zone from Mauritania to Nigeria, and aims to measure, using satellite and topographic data, the distribution and space-time evolution of its major morphological, hydrological, environmental and urban features, so as to provide a reference measure of the current state of the littoral zone and first estimates of its erosion rates. In a first step, the post-doc researcher will develop deep learning architectures allowing the automatic recognition and extraction, from satellite (Sentinel 2, Spot 6-7, Pléiades), airborne (photogrammetry) and topographic data (associated DSMs), of the major features of the littoral area: curvilinear features (such as coastline, hydrographic network, roads, sand dune ridges, sedimentary and geological boundaries, etc.), 3D features (such as buildings, dams, etc.), and textural features (such as vegetation, geological outcrops, water zones, etc.). The experts involved in this proposal will create training data sets, to enable deep learning training. The postdoc will develop algorithms as generic as possible, so that they could be applied to remote data of different types and resolutions, and possibly, to the extraction of other features than littoral structures. In a second step, the developed architectures will be used to automatically map the different littoral features. Time series of satellite images will be acquired on a few target zones, allowing, through repeated extraction of similar features, for the measurement of their space-time dynamics.

Duration: 2 years
Starting date: between September 1st and December 1st, 2019
Salary: gross salary per month 2500 EUR (i.e. approx. 2000 EUR net)

Hosting laboratory: GEOAZUR, Sophia Antipolis (https://geoazur.oca.eu/fr/acc-geoazur)
Other laboratories where the post-doc will work: LUXCARTA (https://luxcarta.com/) and LEGOS (http://www.legos.obs-mip.fr/)

Candidate profile
Strong academic background in applied mathematics, machine learning and image processing. Motivation for academic and/or applied research, and for environmental issues. Curiosity, ability to learn quickly, capacity to work in team and to adapt to different work environments.
The post-doc researcher will need to frequently commute between Géoazur and Luxcarta (~12 km apart), and therefore it would be good to have a driving license and a car.

To apply, please email a full application to Isabelle Manighetti (manighetti@geoazur.unice.fr), Yuliya Tarabalka (ytarabalka@luxcarta.com) and Rafaël Almar (rafael.almar@ird.fr), indicating “IRD_Littoral_post-doc” in the e-mail subject line.
The application should contain:
- A motivation letter demonstrating motivation, strengths from academic or private sectors, and related experience to the position
- CV including publication list
- At least two major publications in pdf
- Minimum 2 reference letters

How to Apply

This job was posted by a member of the IRIS community. For clarification or details about this position, please use the contact information found in the job posting itself.