My thesis title is Sparse Semi-Parametric Recovery of Astronomical Radio-Images. Its goal is to develop new signal processing tools to reconstruct image acquired from radio telescopes. The main topics that I work on are Image Processing, Inverse Problems, Convex Optimization and Deep Learning.
Shape information contained in galaxy images is important in various domains of astrophysics such as Cosmology and Galaxy Evolution. During my PhD, I developed a shape constraint that allows to improve the reconstruction of galaxy images and the recovery of its physical information. As a proof of concept, I developed a sparse recovery algorithm and added the shape constraint to it. For this purpose, I generated a parametric dataset of optical galaxy images. Then I extended the shape constraint to Deep Learning and tested it by generating two realistic dataset containing more than 50,000 images each. The first dataset has optical images and the second one has MeerKAT-like radio images, which allowed us to test the Deep Learning methods we used and improved, for the first time, on radio data.
Conferences & Talks
- CosmoStat Day on Machine Learning in Astrophysics, virtual, France, March 2021 (talk).
- SIAM Conference on Mathematics of Data Science, virtual, USA, August 2020 (talk).
- SKA Cosmology SWG, Paris, France, January 2020 (talk).
- SPIE Wavelets and Sparsity XVIII, San Diego, USA, August 2019 (talk).
- Image Restoration Talk at SDSU Dep. of Math. and Stat., San Diego, USA, August 2019 (speaker).
- Wavelets and Beyond, Orsay, France, June 2019.
- 2018-2020 : Teaching assistant, Signals & Systems at Paris-Diderot university, Paris, France.
- 2019-2020 : Teaching assistant, Circuits & Architecture at Paris-Diderot university, Paris, France.
- Bash and Scripting Tutorial for Researchers, virtual, June 2020
Attended Courses & Workshops
- Sparsity4PSL, Paris, France, June 2019 (summer school).
- Deep Learning in Practice, Gif-sur-Yvette, France, Fall 2019 (MVA Master course).
- Science Writing EDP Sciences, Gif-sur-Yvette, France, February 2019 (workshop).
- Intégrité scientifique dans les métiers de la recherche, Bordeaux, France, October 2018 (MOOC).
Previously to my PhD studies
- Engineering degree, specialized in Image Processing, IMT Atlantique, Brest, France, September 2018.
- SISEA Masters II in Research, specialized in Image Processing, IMT Atlantique, Brest, France, September 2018.
- BS in Mathematics, University of Western Brittany, Brest, France, June 2016.
- Preparatory classes, Math and Physics option, Saint Joseph University, Beirut, Lebanon, June 2015.
I am expected to defend my PhD on the 25th of November, 2021, at CEA Paris-Saclay.
Last updated: October 2021