Current Projects



The scientists of CosmoStat laboratory are actively involved in the Euclid mission. The Euclid mission aims at mapping the geometry of the dark Universe to address questions related to fundamental physics and cosmology on the nature and properties of dark energy, dark matter and gravity, as well as on the physics of the early universe and the initial conditions which seed the formation of cosmic structures. Euclid consists of a 1.2 m aperture telescope with two instruments: the visual imager (VIS) and the near-infrared spectrometer and photometer (NISP). CEA is in charge of several unit of these two instruments. CosmoStat members are highly involved in the Euclid Consortium, they have several responsibilities in the Science Ground Segment and the Science Working Groups.

Euclid will require that we manipulate 3D spherical data. We have developed a 3D spherical harmonic analysis decomposition (Leistedt, Rassat, Refregier, Starck et al, A&A, 2012) and a 3D isotropic wavelet decomposition on the sphere (Lanusse, Starck, Rassat, A&A, 2012). The two related packages were released in October 2011. The software is available from the web page here . We have also taken several responsibility positions in the Euclid consortium (OULE3 lead, and several OULE3 work packages leads).

Other projects are:

  • DEDALE:  Data Learning on Manifolds and Future Challenges. H2020-FETOPEN-2014-2015-RIA, 2015-2018, PI.
  • PHySIS: Sparse Signal Processing Technologies for HyperSpectral Systems, H2020-COMPET-06-2014,  2015-2017, CO-I.
  • SKA: Cosmology Scientific Working Group.
  • COSMIC: Compressed Sensing for Magnetic Resonance Imaging & & Cosmology, 2016-2018.
  • CFIS:  The Canada-France Imaging Survey
  • LENA: (non-LinEar sigNal processing for solving data challenges in Astrophysics, ERC, 2016-2021.

Former Projects

Adv ERC SparseAstro:

This project is supported by the European Research Council grant SparseAstro (ERC-228261) for a period of five years, starting October 1st, 2009. The goals of this project are:

  • Develop tools for analyzing multi-valued astronomical data using sparsity.
  • Create software packages which include these tools.
  • Use these tools in the framework of different astronomical international projects (PLANCK, FERMI, EUCLID).


CS-ORION focus is on the design, testing, and evaluation of compressive sensing (CS) architectures for enhancing the high-quality video acquisition and delivery capabilities of remote sensing devices that will enable them to provide efficient remote imaging in aerial and terrestrial surveillance.


atImages focuses on both sparse and variational methods to extract the complex structures of natural images and textures.


MultID is a ANR-funded project which focuses on sparse models and methods to analyze multispectral data.


We were highly involved in the PLANCK project. We run our tools (component separation, inpainting, ISW detection, non-Gaussianity tests based on sparsity, etc) on PLANCK data. We have also shown on simulations that the CMB lensing effect can be recovered even in the presence of missing data, thanks to sparse representations (Perotto, Bobin et al, A&A, 2010; Plaszczynski et al, A&A, 2012). Finally, we have investigated how the theoretical power spectrum can be estimated from the reconstructed CMB map using sparsity (Paykari, Starck et al, A&A, 2012).


We have developed a new method for denoising and deconvolution of multichannel data on the sphere contaminated with Poisson noise (Schmitt, Starck et al, A&A,2010; Schmitt, Starck et al,, A&A, 2012). A toolbox for Multichannel Spherical Image Debluring has been release in March 2012. The FERMI sources catalog has been derived using our wavelet code (Abdo et al (Starck), ApJS, 2010).


We have realized a study on Compressed Sensing for Herschel (see above). One of our sparsity based algorithms, which decomposes one image into two components, has provided one of the strongest results of Herschel (Andre et al, A&A, 2010). Philippe Andre has presented his results at the Beijing IAU meeting (August, 2012) as a Keynote speaker.


The XMM-LSS catalog has being derived using our wavelet code (Pierre, et al (Starck), MNRAS, 2007).


We are involved in the Canada-France-Hawaii Telescope (CFHT) lensing survey (CFHTLenS). Several analysis (PSF treatment, systematics testing, shape measurement) will be submitted soon, as well as several science publications on cosmological constraints for LCDM models, dark energy and modified gravity. This work is a continuation from publications using previous data releases (Fu, Semboloni, Hoekstra, Kilbinger et al., A&A 2008; Kilbinger et al., A&A 2009). Preliminary results were presented at a special CFHTLenS parallel session and press release at the AAS meeting in Austin, Texas, 2012. For systematics testing and to produce the science results, public software developed by M.Kilbinger (athena, nicaea, cosmo_pmc) has been used.