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:20250330T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20251026T010000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20251002T140000
DTEND;TZID=Europe/Paris:20251002T150000
DTSTAMP:20260419T010136
CREATED:20250929T140707Z
LAST-MODIFIED:20250929T140713Z
UID:11975-1759413600-1759417200@www.greyc.fr
SUMMARY:Séminaire Image : "Anomaly Detection with Pre-Trained Representations"\, Jeremie Pantin
DESCRIPTION:Nous aurons le plaisir d’écouter Jérémie Patin\, MCF CDD au GREYC.\nIl donnera un séminaire IMAGE le jeudi 2 octobre 2025 à 14h en salle de séminaire F-200. \nTitre : « Anomaly Detection with Pre-Trained Representations » \nRésumé : \nIn Anomaly Detection\, relatively few studies have examined how pre-trained models can be used without fine-tuning or weight adaptation\, particularly in multimodal scenarios. We first introduce a lightweight autoencoder approach for identifying textual anomalies\, considering representation from encoder-based large language models. We then extend this perspective to a multimodal setting\, focusing on brain tumor analysis with pre-trained text/vision models. Finally\, we conclude with an xAI problem statement: Can we consider predicted anomalies as post-hoc explainers ?
URL:https://www.greyc.fr/event/seminaire-image-anomaly-detection-with-pre-trained-representations-jeremie-pantin/
LOCATION:ENSICAEN – Batiment F – Salle F-200\, 6 Bd Maréchal Juin\, Caen\, 14050\, France
CATEGORIES:General,Image,News,Seminaire Image
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20251007T104500
DTEND;TZID=Europe/Paris:20251007T114500
DTSTAMP:20260419T010136
CREATED:20250901T144053Z
LAST-MODIFIED:20250924T083853Z
UID:11941-1759833900-1759837500@www.greyc.fr
SUMMARY:Séminaire Algorithmique : Sarah Riva (LIFL\, Univ. Lille) « Control and synthesis of minimal trap spaces in Boolean Network »
DESCRIPTION:Since recent years\, we observe a surge of successful applications of Boolean networks (BNs) in biology and medicine for the modeling and prediction of cellular dynamics in the case of cancer and cellular reprogramming. Such applications face two main challenges: being able to design a qualitative Boolean model which is faithful to the behavior of the biological system and being able to compute predictions to control its (long-term) dynamics. From a computational point of view\, the latter problem mostly depends on the complexity of the dynamical property to enforce\, while the former additionally suffers from the combinatorics of candidate models. \nMinimal trap spaces (MTSs) capture subspaces in which the Boolean dynamics is trapped\, whatever the update mode. They correspond to the attractors of the most permissive mode. Due to their versatility\, the computation of MTSs has recently gained traction\, essentially by focusing on their enumeration. We address the logical reasoning on universal properties of MTSs in the scope of two problems: the reprogramming of Boolean networks for identifying the permanent freeze of Boolean variables that enforce a given property on all the MTSs\, and the synthesis of Boolean networks from universal properties on their MTSs. Both problems reduce to solving the satisfiability of quantified propositional logic formula with 3 levels of quantifiers.
URL:https://www.greyc.fr/event/seminaire-algorithmique-sarah-riva-lifl-univ-lille/
LOCATION:Sciences 3- S3 351
CATEGORIES:Amacc,General,News,Séminaire Algo
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20251009T140000
DTEND;TZID=Europe/Paris:20251009T150000
DTSTAMP:20260419T010136
CREATED:20250929T081731Z
LAST-MODIFIED:20250929T081731Z
UID:11971-1760018400-1760022000@www.greyc.fr
SUMMARY:Séminaire Image : "SIMuLDiTex : Un modèle de diffusion léger\, mono-image et multi-échelle pour la synthèse de texture"\, Pierrick Chatillon
DESCRIPTION:Nous aurons le plaisir d’écouter Pierrick Chatillon\, post-doctorant au GREYC\, Caen.\nIl donnera un séminaire IMAGE le jeudi 9 octobre 2025 à 14h en salle de séminaire F-200. \nTitre : « SIMuLDiTex : Un modèle de diffusion léger\, mono-image et multi-échelle pour la synthèse de texture » \nRésumé : \nNous présentons SIMuLDiTex\, un modèle frugal de diffusion multi-échelle pour la synthèse de textures à partir d’une seule image d’exemple. Cette approche est basée sur une stratégie pyramidale d’échantillonnage et de génération des textures permettant de préserver la fidélité des détails générés\, même à haute résolution. Grâce à un mécanisme original de conditionnement multi-échelle d’un U-Net compact (environ un million de paramètres)\, SIMuLDiTex surpasse les méthodes de l’état de l’art à nombres de paramètres équivalents\, offrant ainsi une solution rapide et de qualité pour la génération de textures de très haute résolution. \nLe séminaire sera précédé d’une intervention d’Othilie Malfilâtre de Normandie Valorisation dans le cadre de sa campagne de la détection au sein de l’équipe Image.
URL:https://www.greyc.fr/event/seminaire-image-simulditex-un-modele-de-diffusion-leger-mono-image-et-multi-echelle-pour-la-synthese-de-texture-pierrick-chatillon/
LOCATION:ENSICAEN – Batiment F – Salle F-200\, 6 Bd Maréchal Juin\, Caen\, 14050\, France
CATEGORIES:General,Image,News,Seminaire Image
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20251014T104500
DTEND;TZID=Europe/Paris:20251014T114500
DTSTAMP:20260419T010136
CREATED:20250901T144223Z
LAST-MODIFIED:20250924T083940Z
UID:11943-1760438700-1760442300@www.greyc.fr
SUMMARY:Séminaire Algorithmique : France Gheeraert (LAMFA\, Univ. Picardie) «String attractors\, ou comment capturer la combinatoire d’un texte»
DESCRIPTION:Les string attractors sont des objets combinatoires introduits par Kempa et Prezza dans le but d’unifier différentes mesures de compressibilité de texte provenant de techniques classiques telles que LZ77 ou la transformée de Burrows-Wheeler. Etant donné un texte\, un string attractor est un ensemble de positions permettant de capturer tous les motifs apparaissant dans ce texte. La recherche d’un tel ensemble de positions est un problème NP-complet. Pour contourner cette difficulté\, les recherches se sont rapidement tournées vers des textes possédant des propriétés combinatoires particulières. \nDans cet exposé\, je présenterai les string attractors et les résultats classiques à leur sujet. J’aborderai ensuite le cas particulier des préfixes de la suite (symbolique) de Fibonacci comme une première étape vers l’étude d’autres textes bien connus en combinatoire des mots.
URL:https://www.greyc.fr/event/seminaire-algorithmique-france-gheeraert-lamfa-univ-picardie/
LOCATION:Sciences 3- S3 351
CATEGORIES:Amacc,General,News,Séminaire Algo
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20251016T140000
DTEND;TZID=Europe/Paris:20251016T150000
DTSTAMP:20260419T010136
CREATED:20250929T083055Z
LAST-MODIFIED:20250929T083055Z
UID:11973-1760623200-1760626800@www.greyc.fr
SUMMARY:Séminaire Image : "Research Aiming for Small-Cost Development of Medical Image Analysis Applications - Insights from a Career Bridging Clinical Practice and Research"\, Mitsutaka Nemoto
DESCRIPTION:Nous aurons le plaisir d’écouter Mitsutaka Nemoto\, Associate Professor of Kindai Univ\, Japan. Il est aussi visiteur au GREYC pour un an.\nIl donnera un séminaire IMAGE le jeudi 16 octobre 2025 à 14h en salle de séminaire F-200. \nTitre : « Research Aiming for Small-Cost Development of Medical Image Analysis Applications – Insights from a Career Bridging Clinical Practice and Research » \nRésumé : \nThis seminar will present the outcomes of my works in medical image analysis AI\, conducted at a university medical school\, a university hospital\, and a training institution for clinical engineers (a paramedical profession in Japan). The central focus or hidden theme of these works is reducing the development costs of the applications.\nKey cost factors in medical image AI development include collecting clinical image data\, acquiring annotation data\, designing application frameworks\, and implementing the systems. With the widespread adoption of (deep) supervised machine learning\, the demand for large\, annotated datasets has grown substantially. However\, many medical institutions in Japan\, particularly those of a middle or small size\, lack the resources to meet these demands. Consequently\, creating and providing customized AI applications trained on data aligned with each hospital’s imaging and diagnostic protocols remains a significant challenge.\nThis seminar will highlight my research aimed at reducing development costs\, such as AI application development using unsupervised learning and self-supervised learning.
URL:https://www.greyc.fr/event/seminaire-image-research-aiming-for-small-cost-development-of-medical-image-analysis-applications-insights-from-a-career-bridging-clinical-practice-and-research-mitsutaka-nemoto/
CATEGORIES:General,Image,News,Seminaire Image
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20251021T104500
DTEND;TZID=Europe/Paris:20251021T114500
DTSTAMP:20260419T010136
CREATED:20250901T144337Z
LAST-MODIFIED:20251017T093914Z
UID:11945-1761043500-1761047100@www.greyc.fr
SUMMARY:Séminaire Algorithmique : Michel Seck (Ecole Politech. Thiès\, Sénégal) « Towards post-quantum Bitcoin blockchain using Dilithium signature »
DESCRIPTION:Bitcoin is one of the famous cryptocurrencies in the world. It is a permissionless blockchain\, and all transactions are stored in a public decentralized ledger. In its security design\, Bitcoin utilizes various cryptographic primitives\, such as hash functions and signature schemes. In the current version of Bitcoin\, the Elliptic Curve Digital Signature Algorithm (ECDSA) is employed\, which is not considered post-quantum secure due to the Shor’s algorithm. In this talk\, we will analyze the potential replacement of ECDSA with Dilithium\, which is a postquantum digital signature based on lattices and recently standardized by NIST as ML-DSA. Bitcoin operates on a pseudonymous system rather than providing complete anonymity. To enhance privacy protection\, the Bitcoin community has adopted a special type of deterministic wallet as outlined in Bitcoin Improvement Proposal 32 (BIP32). We will show how to construct deterministic wallets from Dilithium by first designing DilithiumRK\, a signature scheme with rerandomizable keys build on top of Dilithium. We will then discuss the unlinkability\, unforgeability and efficiency of DilithiumRK and the resulting wallets.
URL:https://www.greyc.fr/event/seminaire-algorithmique-francois-rioult-et-abdelkader-ouali-greyc-caen/
LOCATION:Sciences 3- S3 351
CATEGORIES:Amacc,General,News,Séminaire Algo
END:VEVENT
END:VCALENDAR