Vous êtes ici : Accueil » Kiosque » Annonce

Identification

Identifiant: 
Mot de passe : 

Mot de passe oublié ?
Détails d'identification oubliés ?

Annonce

20 décembre 2019

Internship/PhD : ACDC with deep learning : Automatic Crater Detection and Characterization with deep learning


Catégorie : Stagiaire


Université Paris Saclay is looking for a Internship/PhD to develop a machine learning/deep learning tool to automatically detect craters on spaceborne images.

 

 

Description:
This study takes place in the data deluge from the numerous space missions across the Solar System. The project proposes to develop a tool to automatically detect and characterize the most ubiquitous feature on planetary body : craters.
The aim is to developed a tool to define precise size and position of all craters in the scene, whatever the illumination conditions, the type of sensor and the scale. As a second goal, the project will have to determine the crater characteristics, such primary / secondary (ejecta from a previous impact, not from a direct impactor), presence / absence of rays, erosion level...
This study will take advantage of the machine learning and deep learning libraries available as open source to propose the most versatile and robust detection method. We propose to develop a new tool dedicated to this task. In addition, we propose to organize a worldwide challenge for any researcher/students as an open source strategy, in a framework called RAMP. This platform is designed for collaborative work and gives access to the source code of the participants (not only the results).
Such software pipeline is required to tackle fundamental questions in planetary science to study the surface processes across the Solar System. It will be a crucial tool to precisely date the surface and open a new era for onboard decisions on landing or targeting, to maximize the science return of future deep space missions.

Candidate requirements :
The candidate must have a engineer or master grade in machine learning/data mining or in planetary science. Double competence in both fields will be encouraged. An excellent level of programming skills is required (Python, linux). We expect the candidate to have a good level of communication in English (written and oral).
Please send following pieces :
- CV
- all marks in superior education

Duration :
Internship : 6 months max, in spring/summer 2020
PhD : 3 years, starting from October 2020

Location :
UMR8148 GEOPS
Bât 509, Université Paris Saclay
91405 ORSAY, FRANCE

Contact :
Frédéric Schmidt, Professor, GEOPS, Univ. Paris Saclay, expert of Planetary Science and surface processes, ​http://planeto.geol.u-psud.fr/spip.php?article83
Hugues Talbot, Professor, Ecole CentraleSupelec, Paris Saclay, expert in Artificial Intelligence and computer vision, ​https://hugues-talbot.github.io/

 

Dans cette rubrique

(c) GdR 720 ISIS - CNRS - 2011-2020.