Objectives: Develop, optimize, and implement state-of-the-art object detection and segmentation Deep Learning algorithms for the task of feature detection using remote and/or in-situ sensing of rocky planetary bodies such as Mars, The Moon, and Mercury. (Credit: NASA/JPL/University of Arizona - HiRISE) Supervisor: Prof. Nigel Mason
Project Summary
The era of Big Data is transforming the way scientists approach their research, and ultimately how science progresses and discoveries are made. Current space missions have highlighted a critical need for the development of new tools capable of processing and analysing Big Data from space. In parallel, recent missions, benefiting from technological progresses, have raised the level of collected data to unprecedented levels. Advances in observation systems and instruments require equal advances in data management and analysis. In this framework, Artificial Intelligence (AI) is a powerful tool that is becoming more common across a wide range of fields, including planetary sciences and Earth Observations.
The focus of this project is on mapping planetary surfaces with A.I. The distribution and morphology of diverse surface features on Solar system planets can shed light on their origins, and on the underlying processes of geological activity. Automatic object detection and classification are also relevant for human and robotic exploration. The goal of this project is to develop state-of-the-art object detection and segmentation Deep Learning algorithms for feature detection using remote and/or on-ground sensing observations of rocky planetary bodies such as Mars, The Moon, and Mercury. Major challenges are to use multi-resolution and multi-spectral data, as well as to combine information from different instruments (data fusion).
This PhD project is part of a collaboration between Kent University and ACRI-ST. ACRI-ST (https://www.acri-st.fr/ and https://astro.acri-st.fr/) is an SME of the space sector, bringing together scientific research and engineering of complex Information and communications technology (ICT) systems. ACRI-ST is supplier of space agencies and provides engineering services for satellite data, addressing the whole chain of the Space activities, from support to on-board instrument specifications and data ground segment development to operational services for end-users. ACRI-ST has both technical (IT) and scientific competences in Space Sciences and Earth Observations, including expertise in planetary science, astrophysics, data processing, machine learning, and development of (scientific) exploitation platforms. The candidate should expect to spend significant part of his/her PhD in our establishments in France (Grasse, Toulouse).
We look for an inquisitive candidate with a background in computer science or astrophysics. The student should be skilled in computing coding (preferably Python) and familiarity with AI techniques and planetary geology is a plus.
Sub-topics
- Remote and in-situ imaging & sensing
- Object detection and image segmentation with deep learning
- Study of planetary geological surface features
Profile PhD student
The student should have an MSc in computer science or (astro)physics and be familiar with some of the following:
- Coding (python)
- Deep Learning
- Planetary (Moon/Mars/Mercury) geology
Skills
- Inquisitive and critical thinker
- Independent mindset
- Strong communication skills
- Fluent in English
For further details and discussion contact Prof. Nigel Mason (n.j.mason[at]kent.ac.uk).