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:20230326T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20231029T010000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20230202T140000
DTEND;TZID=Europe/Paris:20230202T153000
DTSTAMP:20260423T135413
CREATED:20230201T114800Z
LAST-MODIFIED:20230307T083457Z
UID:11087-1675346400-1675351800@www.greyc.fr
SUMMARY:Séminaire IMAGE : « Patch and attention for image editing » (Nicolas Cherel)
DESCRIPTION:Nous aurons le plaisir d’écouter Nicolas Cherel\, Doctorant à Télécom Paris (Institut Polytechnique de Paris).\nIl donnera un séminaire IMAGE\, le jeudi 2 février 2023 à 14h00\, en salle de séminaire F-200.\nTitre: « Patch and attention for image editing »\nRésumé : We show through two different examples that patch-based methods remain relevant despite the widespread use of neural networks for many image editing tasks.\nWe first present a patch-based algorithm for single image generation that performs as well as a neural network without requiring a costly training phase. We ensure visual fidelity and diversity of the results by carefully choosing the initialization of the algorithm.\nIn the second part\, we show that patch-based algorithms can benefit to modern techniques such as attention mechanisms. The use of attention has helped deep learning introduce long range dependencies but computing the full attention matrix is an expensive step with heavy memory and computational loads. We propose an efficient attention layer based on the stochastic algorithm PatchMatch\, which is used for determining approximate nearest neighbors. Our layer has a greatly reduced memory complexity compared to other attention layers\, scaling to high resolution images. \n  \nVenez nombreux!
URL:https://www.greyc.fr/event/seminaire-image-patch-and-attention-for-image-editing-nicolas-cherel/
LOCATION:ENSICAEN – Batiment F – Salle F-200\, 6 Bd Maréchal Juin\, Caen\, 14050\, France
CATEGORIES:General,Image,Seminaire Image
END:VEVENT
BEGIN:VEVENT
DTSTART;VALUE=DATE:20230207
DTEND;VALUE=DATE:20230208
DTSTAMP:20260423T135413
CREATED:20230125T163951Z
LAST-MODIFIED:20230125T163951Z
UID:11078-1675728000-1675814399@www.greyc.fr
SUMMARY:Journée de la fédération NormaSTIC
DESCRIPTION:Venez profiter d’un programme qui inclut des exposés d’invités reconnus et des exposés relatifs à nos activités en Normandie et liées à la vie de la fédération. Parmi les exposés invités\, nous avons une décodeuse du numérique et une médaille de bronze du CNRS. \nLa journée aura lieu au Wip (à Colombelles) https://www.le-wip.com/ \nVoir le programme à https://www.normastic.fr/event/journee-de-la-federation-2023/ \n  \n 
URL:https://www.greyc.fr/event/journee-de-la-federation-normastic/
LOCATION:WIP\, Colombelles\, Rue des Ateliers 14460\,\, Colombelles\, 14460\, France
CATEGORIES:General
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20230221T100000
DTEND;TZID=Europe/Paris:20230221T110000
DTSTAMP:20260423T135413
CREATED:20230125T164357Z
LAST-MODIFIED:20230213T105441Z
UID:11083-1676973600-1676977200@www.greyc.fr
SUMMARY:Séminaire Algorithmique : Mehdi Naima « Extending Brandes algorithm to improve betweenness centrality computation in temporal graphs with discrete and continuous time »
DESCRIPTION:Betweenness centrality has been a long subject of study in network science since it was introduced by Freeman in 1977. This centrality measure assesses the importance of nodes in a graph\, it has been used for example in social\, biological and research collaboration networks. Moreover\, betweenness centrality has been used in graph partitioning and community detection in the well-known Girvan-Newan algorithm.\n\nThis centrality measure is based on the enumeration of shortest paths passing through a node. A simple approach to compute betweenness centrality for all the nodes of a static graph is to use Floyd-Warshall algorithm that runs in O(n3). Brandes in 2001 published an algorithm that runs in O(nm + n2 log n) on weighted graphs\,  it is still considered one of the best theoretical results on the question.\n\nBetweenness centrality has also been extended to temporal graphs. Temporal graphs have edges that bear labels according to the time of the interactions between the nodes. Betweenness centrality has been extended to the temporal graph settings\, and the notion of paths has been extended to temporal paths. We will see that we are able to deploy Brandes algorithm to its full extent and improve the running time of recent results to O(nmT + n2Tlog(nT)). We will also discuss how Brandes algorithm can also be generalized to stream graphs which are dynamic graphs with continuous time and dynamicity on the nodes.
URL:https://www.greyc.fr/event/seminaire-algorithmique-mehdi-naima/
LOCATION:Sciences 3- S3 351
CATEGORIES:Amacc,General,Séminaire Algo
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20230228T100000
DTEND;TZID=Europe/Paris:20230228T120000
DTSTAMP:20260423T135413
CREATED:20230125T165233Z
LAST-MODIFIED:20230224T142635Z
UID:11085-1677578400-1677585600@www.greyc.fr
SUMMARY:Séminaire Algorithmique : « Algorithmes pour la Dimension Métrique dans les graphes dirigés » Antoine Dailly (LIMOS\, Univ. Clermont-Ferrand)
DESCRIPTION:Résumé :\nLe problème de la Dimension Métrique d’un graphe se pose de la façon suivante : on cherche un ensemble R de sommets de taille minimale tel que\, pour toute paire de sommets du graphe\, il existe un sommet de R dont les distances aux deux sommets de la paire sont distinctes. Ce problème a été principalement étudié dans les graphes non-dirigés\, et il a attiré l’attention ces dernières années\, notamment en raison de sa difficulté : il est NP-complet et son meilleur facteur d’approximation en temps polynomial est log(n)\, y compris sur des classes de graphes très restreintes. \nNous considérons ce problèmes dans les graphes dirigés et orientés (la différence est que les graphes dirigés peuvent contenir des 2-cycles\, au contraire des graphes orientés)\, pour lesquels la Dimension Métrique a été récemment étudiée. Nous montrons que le problème reste NP-dur\, même dans les graphes orientés planaires bipartis de degré maximum 6. Cependant\, nous donnons des algorithmes linéaires résolvant le problème sur les arbres dirigés (les graphes dirigés dont le graphe sous-jacent est un arbre) et les orientations de graphes unicycliques. Enfin\, nous avons un algorithme paramétré par la largeur modulaire. \n\nAbstract:\nIn graph theory\, the Metric Dimension problem is the following: we are looking for a minimum-size set R of vertices\, such that for any pair of vertices of the graph\, there is a vertex from R whose two distances to the vertices of the pair are distinct. This problem has mainly been studied on undirected graphs\, and has gained a lot of attention in recent years\, mainly because of its difficulty: it is NP-complete and has a best polynomial-time approximation factor of log(n) even on very restricted graph classes. \nWe focus our study on directed and oriented graphs (the difference is that directed graphs may contain 2-cycles\, unlike oriented graphs)\, for which the Metric Dimension has been recently studied. We prove that Metric Dimension remains NP-hard\, even on planar bipartite oriented graphs of maximum degree 6. However\, we find linear-time algorithms solving the problem on directed trees (directed graphs whose underlying graph is a tree) and on orientations of unicyclic graphs. Finally\, we give a fixed-parameter-tractable algorithm for directed graphs when parameterised by modular-width.
URL:https://www.greyc.fr/event/seminaire-algorithmique-antoine-dailly-limos-univ-clermont-ferrand/
LOCATION:Sciences 3- S3 351
CATEGORIES:Amacc,General,Séminaire Algo
END:VEVENT
END:VCALENDAR