BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CRAL - ECPv6.15.20//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:CRAL
X-ORIGINAL-URL:https://cral.osu-lyon.fr
X-WR-CALDESC:Évènements pour CRAL
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:Europe/Paris
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20240331T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20241027T010000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20250330T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20251026T010000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20260329T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20261025T010000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20250120T140000
DTEND;TZID=Europe/Paris:20250120T170000
DTSTAMP:20260510T064044
CREATED:20250113T174832Z
LAST-MODIFIED:20250113T175025Z
UID:7681-1737381600-1737392400@cral.osu-lyon.fr
SUMMARY:CRAL@Obs PhD Defense by Théo Santos (CRAL\, LGL)
DESCRIPTION:Generative neural networks for parametrization in inverse methods: applications in astrophysics and geophysics \nIn physical science\, it is common to study the characteristics of an object that are not directly available but which certain effects within the physical system are observable. The observations can then be used to deduce the properties of the object of interest: the approach known as “inverse” focuses on such problems. An essential question is the parametrization: how to properly define the parameters that define the object to study? The choice is crucial\, as it determines the formulation of the inverse problem and therefore affects its difficulty to solve. \nIn recent years\, machine learning\, particularly generative neural networks\, has offered a promising opportunity for addressing parametrization. These generative networks can generate\, at will\, objects similar to training objects\, from only a few input parameters. A generative network can thus be trained on a set of possible objects and then used for parametrization. This approach presents several advantages: it is computationally efficient\, it attenuates the irregularities between the parameters defining the object and the observations\, it constrains the problem to a class of models similar to those in the training set\, and it requires few degrees of freedom. \nThis thesis is devoted to the methodological analysis of this parametrization. In particular\, I propose various tools to ensure the optimal training of a generator for this parametrization. I highlight the benefits\, limitations\, and specific considerations of this approach. To illustrate this methodological study\, I rely on a toy problem consisting of recovering a clear image of a ring from a blurred and noisy version. I use it to illustrate the various tools and issues related to parametrization\, which I apply within sampling and optimization inverse methods. Subsequently\, I apply this approach to a geophysical problem and an astrophysical one to demonstrate its relevance in these domains. In geophysics\, I apply this parametrization to a “downscaling” problem: the objective is to use a smooth image of the Earth’s mantle to derive a finer image containing small-scale details. To achieve this\, I train the generative network on models created through geodynamic simulations; the network acts as an approximation of the simulation\, more actionable for inverse problems. This application\, performed on synthetic data\, demonstrates the potential of this parametrization for future applications with real data. In astrophysics\, I apply this approach to the generation of point spread functions (PSF). In astrophysical imaging\, the PSF is an object\, characteristic of the optical system\, that defines how an image is blurred; estimating it accurately allows for efficient deblurring of an image. I thus parameterize the PSFs with a generative network\, making the generator as an useful tool in inverse problems involving the PSF. For this case\, the parametrization is performed on real PSF images\, enabling to explore specific issues related to using real data. \n— \nThe presentation will be done in French.
URL:https://cral.osu-lyon.fr/event/cralobs-phd-defense-by-theo-santos-cral-lgl/
LOCATION:Observatoire
ATTACH;FMTTYPE=image/png:https://cral.osu-lyon.fr/wp-content/uploads/2023/11/image_defaut.png
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20250121T110000
DTEND;TZID=Europe/Paris:20250121T110000
DTSTAMP:20260510T064044
CREATED:20241218T110645Z
LAST-MODIFIED:20250113T171322Z
UID:7645-1737457200-1737457200@cral.osu-lyon.fr
SUMMARY:CRAL@Obs seminar by Nicolas Dobigeon (IRIT)
DESCRIPTION:Multiband image fusion under spectrally varying spatial blurs — Application to JWST data \nMultiband image fusion has been extensively studied in the literature of Earth observation. This task aims at recovering a multiband image of high spatial and high spectral resolutions from complementary measurements of lower spatial or spectral resolutions. While several methods developed in this context are able to fuse multiband images efficiently\, they are not suitable to handle astronomical images\, in particular in the context of observations made by the James Webb Space Telescope (JWST). Beyond the high dimensionality of the data\, one major challenge lies in the operators associated with spatial blurs that should be considered as spectrally varying. This crucial issue significantly increases the complexity of the forward models and makes the state-of-the-art fusion methods inoperative. The main contributions reported in this talk tackle both challenges: a new framework is proposed to fuse large-scale astronomical data efficiently while taking into account the specificities of astronomical imaging.
URL:https://cral.osu-lyon.fr/event/cralobs-seminar-by-nicolas-dobigeon-irit/
LOCATION:Observatoire
ATTACH;FMTTYPE=image/png:https://cral.osu-lyon.fr/wp-content/uploads/2023/11/image_defaut.png
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20250131T110000
DTEND;TZID=Europe/Paris:20250131T120000
DTSTAMP:20260510T064044
CREATED:20250124T164739Z
LAST-MODIFIED:20250129T101537Z
UID:7720-1738321200-1738324800@cral.osu-lyon.fr
SUMMARY:CRAL@ENS internal seminar by Armand Leclerc
DESCRIPTION:New tools from wave topology for asteroseismology\nAsteroseismology studies waves in stars and uses them to infer what their internal structure is. Today\, hundreds of stars of every kind show oscillations which allowed us to deduce their mass\, radius\, rotation rate\, width of convective zone\, etc. Wave topology proposes a new light under which to look at the wave equation\, in order to obtain theoretical characterization of large-scale oscillation modes. This framework comes from quantum physics\, and complement the classical theories of stellar oscillations. I will present its concepts\, and show two new results obtained by it : a characterization of the fundamental mode n=0 at long horizontal wavelength\, and a correction term for low n modes at short horizontal wavelength. Both these predictions are confirmed in the case of the Sun\, opening perspectives for analytical low n modes seismology in other stars. \nWhere?\nENS-Monod – Amphi F
URL:https://cral.osu-lyon.fr/event/cralobs-internal-seminar-by-armand-leclerc/
LOCATION:ENS
ATTACH;FMTTYPE=image/png:https://cral.osu-lyon.fr/wp-content/uploads/2023/11/image_defaut.png
END:VEVENT
END:VCALENDAR