BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//GREYC UMR CNRS 6072 - Groupe de Recherche en Informatique, Image, et Instrumentation de Caen - ECPv5.7.0//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:GREYC UMR CNRS 6072 - Groupe de Recherche en Informatique, Image, et Instrumentation de Caen
X-ORIGINAL-URL:https://www.greyc.fr
X-WR-CALDESC:évènements pour GREYC UMR CNRS 6072 - Groupe de Recherche en Informatique, Image, et Instrumentation de Caen
BEGIN:VTIMEZONE
TZID:Europe/Paris
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:20260505T104500
DTEND;TZID=Europe/Paris:20260505T114500
DTSTAMP:20260615T145243
CREATED:20260526T120349Z
LAST-MODIFIED:20260526T120349Z
UID:12123-1777977900-1777981500@www.greyc.fr
SUMMARY:Séminaire Algorithmique : « From Elastic Degenerate Strings to Full-Text Indexing: Suffix Sorting\, BWT\, and FM-Index Construction and Search »\, Francesco Pio Marino (Univ. Catane\, Italie)
DESCRIPTION:Elastic degenerate strings provide a flexible framework for representing sequences with structured variability\, generalizing classical string models while preserving algorithmic tractability. In this talk\, we present a comprehensive framework for indexing such objects\, culminating in the construction of an FM-index that supports efficient pattern matching. \nWe begin by revisiting the notion of elastic degeneracy and its algorithmic implications. We then describe how classical building blocks—suffix sorting\, the Burrows–Wheeler Transform (BWT)\, and wavelet trees—can be extended to this richer setting. In particular\, we outline a linear-time suffix sorting approach based on an adaptation of the DC3 algorithm\, followed by the construction of the BWT and the associated wavelet tree representation. \nFinally\, we show how these components combine into a full FM-index and describe how the classical backward search procedure can be adapted to this setting\, enabling efficient pattern matching over elastic degenerate strings. The resulting framework bridges combinatorial pattern matching and compressed indexing\, opening the way to scalable querying in settings where uncertainty and variability are intrinsic.
URL:https://www.greyc.fr/event/seminaire-algorithmique-from-elastic-degenerate-strings-to-full-text-indexing-suffix-sorting-bwt-and-fm-index-construction-and-search-francesco-pio-marino-univ-catane-italie/
LOCATION:Sciences 3- S3 351
CATEGORIES:Amacc,General,News,Séminaire Algo
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20260505T110000
DTEND;TZID=Europe/Paris:20260505T120000
DTSTAMP:20260615T145243
CREATED:20260512T080924Z
LAST-MODIFIED:20260512T080924Z
UID:12097-1777978800-1777982400@www.greyc.fr
SUMMARY:Séminaire Image: Kirchhoff Forests: Theory and Applications par Yusuf Yigit Pilavci
DESCRIPTION:Nous aurons le plaisir d’écouter Yusuf Yigit Pilavci \, Postdoc a l’Institut Mines Telecom Nord Europe.\nIl donnera un séminaire IMAGE le mardi 05 mai 2026 à 11h en salle de séminaire F-200. \nTitre : « Kirchhoff Forests: Theory and Applications » \nRésumé : Many machine learning problems involve data structured as graphs\, from social networks and recommendation systems to traffic and biological networks. A key tool for analyzing these graphs is the graph Laplacian\, which captures important structural information but becomes expensive to compute at large scale.\nIn this presentation\, I will introduce a probabilistic tool called Kirchhoff forests for accelerating expensive operations involving graph Laplacians\, such as computing their inverse\, estimating the trace of the inverse\, and estimating effective resistances. This probabilistic method is based on a random process that allows us to generate graph sketches strongly connected to the graph Laplacian through various theoretical guarantees. In turn\, this approach leads to efficient algorithms for applications in graph signal processing and machine learning. \n 
URL:https://www.greyc.fr/event/seminaire-image-kirchhoff-forests-theory-and-applications-par-yusuf-yigit-pilavci/
LOCATION:ENSICAEN – Batiment F – Salle F-200\, 6 Bd Maréchal Juin\, Caen\, 14050\, France
CATEGORIES:General,Image,Seminaire Image
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20260512T104500
DTEND;TZID=Europe/Paris:20260512T114500
DTSTAMP:20260615T145243
CREATED:20260526T120536Z
LAST-MODIFIED:20260526T120536Z
UID:12125-1778582700-1778586300@www.greyc.fr
SUMMARY:Séminaire Algorithmique : « Un automate pour les caractériser tous »\, Théo Grente (GREYC)
DESCRIPTION:Au début des années 2000\, Okhotin a introduit deux familles de grammaires formelles\, les grammaires conjonctives et les grammaires booléennes\, qu’il présente comme “le véritable cas général des grammaires sans contexte”. Ces grammaires enrichissent les grammaires algébriques par l’ajout d’une opération de conjonction pour les grammaires conjonctives et de la négation en plus de la conjonction pour les grammaires booléennes. Ces deux nouvelles grammaires (et leurs restrictions linéaires) viennent ainsi étoffer la zoologie des classes de langages formels sans contexte. L’un des critères permettant de mesurer l’importance d’une classe de langages est qu’elle dispose d’une définition équivalente par une famille d’automates. On peut ainsi citer la caractérisation des langages réguliers par les automates finis\, celle des algébriques par les automates à pile ou encore la caractérisation de la restriction linéaire des grammaires conjonctives par les automates treillis. Ces caractérisations nous permettent de mieux comprendre le pouvoir d’expression des grammaires mais ne facilitent pas forcément leur comparaison. \nDans cet exposé\, je présenterai une famille d’automates\, appelés automates SCYK\, permettant d’obtenir une caractérisation uniforme des classes de langages citées. Cette famille d’automates est inspirée de l’algorithme classique pour la reconnaissance de langages algébriques découvert indépendamment par Sakai\, Cocke\, Younger et Kasami (d’où son nom). Dans leur version la plus générale\, ces automates caractérisent exactement les langages booléens puis\, en leur ajoutant différentes restrictions\, ceux-ci nous permettent de caractériser naturellement les langages conjonctifs\, algébriques (avec leurs restrictions linéaires) ainsi que les langages réguliers.
URL:https://www.greyc.fr/event/seminaire-algorithmique-un-automate-pour-les-caracteriser-tous-theo-grente-greyc/
LOCATION:Sciences 3- S3 351
CATEGORIES:Amacc,General,News,Séminaire Algo
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20260519T104500
DTEND;TZID=Europe/Paris:20260519T114500
DTSTAMP:20260615T145243
CREATED:20260526T112223Z
LAST-MODIFIED:20260526T112223Z
UID:12114-1779187500-1779191100@www.greyc.fr
SUMMARY:Séminaire Algorithmique : « Effective Asymptotics of Combinatorial Systems »\, Carine Pivoteau (LIGM\, Univ. Paris-Est Marne la Vallée)
DESCRIPTION:In their book “Analytic Combinatorics”\, Flajolet and Sedgewick describe a general approach that starts from a combinatorial description\, translates this description into equations satisfied by generating functions\, views these generating functions as analytic functions and exploits their singular behavior to deduce asymptotic properties of the combinatorial objects when their size becomes large. \nWith Bruno Salvy\, we developed computational tools that automate large parts of this approach and in this talk I will outline the main steps.
URL:https://www.greyc.fr/event/seminaire-algorithmique-effective-asymptotics-of-combinatorial-systems-carine-pivoteau-ligm-univ-paris-est-marne-la-vallee/
LOCATION:Sciences 3- S3 351
CATEGORIES:Amacc,General,News,Séminaire Algo
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20260521T140000
DTEND;TZID=Europe/Paris:20260521T153000
DTSTAMP:20260615T145243
CREATED:20260429T133641Z
LAST-MODIFIED:20260505T085221Z
UID:12089-1779372000-1779377400@www.greyc.fr
SUMMARY:Séminaire Image: Shouhei Hanaoka et Rie Tanaka
DESCRIPTION:Nous aurons le plaisir d’écouter Shouhei Hanaoka\, Associate Professor (Graduate School of Medicine\, The University of Tokyo) et Rie Tanaka (Pharmaceutical and Health Sciences\, Kanazawa University).\nIl donneront un séminaire IMAGE le jeudi 21 mai 2026 à 14h en salle de séminaire F-200. \nTitre (Shouhei Hanaoka) : « Clinical importance of trees and graphs: An example of HoTPiG for cerebral aneurysm detection » \nRésumé :  \nThe human body has many tree-like structures such as blood vessels\, bronchi\, bile duct and nervous system.  In this talk\, I am going to present my old work for detecting local abnormality in such tree-like structures.\nThe proposed method was named as HoTPiG (Histogram of Triangular Paths in Graph).\nGiven a graph structure extracted from a binarized volume\, the proposed feature extraction algorithm can effectively encode both the morphological characteristics and the local branching pattern of the structure around each graph node (e.g.\, each voxel in the vessel).\nThe features are derived from a 3-D histogram whose bins represent a triplet of shortest path distances between the target node and all possible node pairs near the target node.\nThe extracted feature set is a vector with a fixed length and is readily applicable to state-of-the-art machine learning methods.\nI will show some examples of lesion detections in the lung and the cerebral vessel.\n\nTitre (Rie Tanaka) : « From Static X-ray to Functional Imaging: AI-driven Advances in Thoracic Radiography » \nRésumé :  \nChest radiography is the most widely used imaging modality for screening and follow-up. Dynamic chest radiography (DCR) extends conventional radiography by enabling functional assessment using flat-panel detector (FPD)-based systems. Quantitative and time-series analysis of DCR allows evaluation of respiratory and circulatory dynamics through lung density changes\, diaphragm motion\, and tracheal diameter. However\, these approaches remain limited by the two-dimensional projection nature of X-ray imaging\, motivating the integration of artificial intelligence and in silico–based strategies to enable higher-dimensional understanding of spatiotemporal patterns.\nWe present a unified framework integrating DCR\, artificial intelligence\, and in silico–based training strategies. We first enhance two-dimensional projections via image decomposition\, then estimate lung volume and respiratory function\, and finally reconstruct four-dimensional (4D) representations from dynamic X-ray time-series data using deep learning. Developed using large-scale virtual datasets and validated on clinical cases\, these approaches may enable functional and quantitative chest radiography\, bridging the gap between low-cost X-ray imaging and advanced modalities such as CT.
URL:https://www.greyc.fr/event/seminaire-image-shouhei-hanaoka-et-rie-tanaka/
LOCATION:ENSICAEN – Batiment F – Salle F-200\, 6 Bd Maréchal Juin\, Caen\, 14050\, France
CATEGORIES:General,Image,Seminaire Image
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20260526T104500
DTEND;TZID=Europe/Paris:20260526T114500
DTSTAMP:20260615T145243
CREATED:20260526T112049Z
LAST-MODIFIED:20260526T112049Z
UID:12112-1779792300-1779795900@www.greyc.fr
SUMMARY:Séminaire Algorithmique : « Weihrauch et la topologie »\, Emmanuel Rauzy (LACL\, Univ. Paris-Est Créteil)
DESCRIPTION:Le but de cet exposé est d’introduire les deux fondements de l’analyse calculable : la théorie des espaces représentés et ses liens avec la topologie\, et la réduction de Weihrauch. \nJ’insisterai en particulier sur le fait que la réduction de Weihrauch est définie grâce à une notion de multifonction continue que l’on ne peut définir que dans la catégorie des espaces représentés. Enfin\, je présenterai de nouveaux outils\, développés avec Vasco Brattka\, pour déterminer facilement la topologie finale d’une représentation.
URL:https://www.greyc.fr/event/seminaire-algorithmique-weihrauch-et-la-topologie-emmanuel-rauzy-lacl-univ-paris-est-creteil/
LOCATION:Sciences 3- S3 351
CATEGORIES:Amacc,General,News,Séminaire Algo
END:VEVENT
END:VCALENDAR