- Annonce

[Seminaire-LRDE] Séminaire Performance et Généricité du LRDE le Mercredi 15 juin 2011
by Jonathan Fabrizio 08 Jun '11

08 Jun '11

Chers collègues, La prochaine session du séminaire Performance et Généricité du LRDE (Laboratoire de Recherche et Développement de l'EPITA) aura lieu le Mercredi 15 juin 2011 (14h-17h). Au programme: * 14h: Modèle basé-contexte pour l'annotation automatique du multimédia -- Nicolas Ballas http://www.kalisteo.fr/en/index.htm http://caor.mines-paristech.fr/ Ces dernières années ont vu exploser le nombre de vidéos disponibles sur internet. Pour permettre leur exploitation, il est nécessaire de mettre en place des systèmes analysant automatiquement ces données multimédia. De tels systèmes permettent notamment d'indexer automatiquement des vidéos en fonction de leurs contenus. Durant cette présentation, je m'intéresserai aux récentes avancées effectuées dans ce domaine. Je présenterai des descripteurs vidéos, développés dans le cadre de ma thèse, qui capturent le mouvement et l'apparence d'une vidéo pour les résumer dans une courte signature. Ces signatures peuvent être utilisées a posteriori pour détecter différents évènements ou concepts dans les vidéos. -- Nicolas Ballas effectue sa thèse entre le laboratoire LVIC du CEA/List et le laboratoire CAOR de l'école des Mines de Paris. Il s'intéresse principalement au problème de perception automatique à travers l'étude d'algorithmes combinant vision par ordinateur et apprentissage automatique. * 15h30: Traitement d'images sur processeur graphique avec CUDA et C++ -- Matthieu Garrigues http://uei.ensta-paristech.fr Conçus à l'origine pour le rendu 2D et 3D, les processeurs graphiques (GPU) peuvent aujourd'hui être considérés comme des processeurs génériques massivement parallèles. Mais ce parallélisme impose des contraintes sur les algorithmes implantés et les types de données utilisés. D'autre part, le bus de communication entre le processeur central (CPU) et le GPU peut être un goulot d'étranglement. Ce séminaire débutera par un aperçu des avantages et inconvénients de la programmation GPU, puis je présenterai l'implantation d'un algorithme temps réel de suivi de points dans une vidéo. Je terminerai par l’introduction de deux petites boîtes à outils : Cuimg et Dige. Cuimg utilise C++ pour simplifier l'écriture d'algorithmes de traitement d'images avec CUDA, tandis que Dige, basée sur le framework Qt, permet le prototypage rapide d'interfaces graphiques. -- Matthieu Garrigues est diplômé de la promotion CSI 2009 de l'EPITA. Depuis, il s'intéresse au développement et l'implantation d'applications de vision par ordinateur sur des architectures parallèles. Il est actuellement ingénieur de recherche à l'unité d'électronique et d'informatique de l'ENSTA. Pour plus de renseignements, consultez http://seminaire.lrde.epita.fr/. L'entrée du séminaire est libre. Merci de bien vouloir diffuser cette information le plus largement possible. _______________________________________________ Seminaire mailing list Seminaire(a)lrde.epita.fr https://www.lrde.epita.fr/mailman/listinfo/seminaire

1 0

Publication: SUMo'11: Generalized Büchi Automata versus Testing Automata for Model Checking
by Alexandre Duret-Lutz 06 Jun '11

06 Jun '11

We are happy to announce that the following paper has been accepted for publication at the 2nd International Workshop on Scalable and Usable Model Checking for Petri Nets and other models of Concurrency (SUMo'11) that will take place in Newcastle, UK, on June 21, 2011. Generalized Büchi Automata versus Testing Automata for Model Checking A.-E. Ben Salem (1&2) A. Duret-Lutz (1) F. Kordon (2) (1) Laboratoire de Recherche et Développement de l'Epita (LRDE) (2) Laboratoire d'Informatique de Paris 6 (LIP6) http://publis.lrde.epita.fr/201106-SUMO Geldenhuys and Hansen have shown that a kind of omega-automaton known as testing automata can outperform the Büchi automata traditionally used in the automata-theoretic approach to model checking. This work completes their experiments by including a comparison with generalized Büchi automata; by using larger state spaces derived from Petri nets; and by distinguishing violated formulae (for which testing automata fare better) from verified formulae (where testing automata are hindered by their two-pass emptiness check). -- Alexandre Duret-Lutz

1 0

Publication: ICDAR'11: A Free Software Framework for Document Image Analysis
by Guillaume Lazzara 06 Jun '11

06 Jun '11

We are happy to announce that the following paper has been accepted for publication at the 11th International Conference on Document Analysis and Recognition (ICDAR) that will take place at Beijing, China, on September 18 - 21, 2011: Guillaume Lazzara (1), Roland Levillain (1), Thierry Géraud (1), Yann Jacquelet (1), Julien Marquegnies (1), Arthur Crépin-Leblond (1) A Free Software Framework for Document Image Analysis http://publis.lrde.epita.fr/201109-ICDAR (1) EPITA Research and Development Laboratory (LRDE) Electronic documents are being more and more usable tanks to better and more affordable network, storage and computational equipment. But in order to benefit from computer-aided document management, paper documents must be digitized and analyzed. This task may be challenging at several levels. Data may be of multiple types thus requiring different adapted processing chains. The tools to be developed should also take into account the needs and knowledge of users, ranging from a simple graphical application to a complete programming framework. Finally, the data sets to process may be large. In this paper, we expose a set of features that a Document Image Analysis framework should provide to address the previous issues. These ideas are implemented as an open source module built on top of a generic and efficient image processing platform. Our solution features services such as preprocessing filters, text detection, page segmentation and document reconstruction (as XML, PDF or HTML documents). This framework, composed of reusable software components, can be used to create full-fledged graphical applications, small utilities, or processing chains to be integrated into third-party projects.

1 0

paper accepted INTERSPEECH 2011
by Reda Dehak 01 Jun '11

01 Jun '11

I'm happy to inform you that our manuscript Paper ID: 95209 Title: Language Recognition via Ivectors and Dimensionality Reduction has been accepted for oral presentation at the INTERSPEECH conference in session ID 2120 "Language identification". -- Réda DEHAK tel : +33 (0)1 44 08 01 86 mailto : reda.dehak(a)lrde.epita.fr LRDE - EPITA 14-16, rue Voltaire F-94270 Le Kremlin Bicêtre cedex France

1 0

[CfP] ACM Symposium on Applied Computing: Separation of Concerns
by Didier Verna 30 May '11

30 May '11

Programming for Separation of Concerns (PSC) at ACM Symposium on Applied Computing (SAC) March 25-29, 2012 Riva del Garda (Trento) Italy Description and Objectives -------------------------- Complex systems are intrinsically expensive to develop because several concerns must be addressed simultaneously. Once the development phase is over, these systems are often hard to reuse and evolve because their concerns are intertwined and making apparently small changes force programmers to modify many parts. Moreover, legacy systems are difficult to evolve due to additional problems, including: lack of a well defined architecture, use of several programming languages and paradigms, etc. Separation of concerns (SoC) techniques such as computational reflection, aspect-oriented programming and subject-oriented programming have been successfully employed to produce systems whose concerns are well separated, thereby facilitating reuse and evolution of system components or systems as a whole. However, a criticism of techniques such as computational reflection is that they may bring about degraded performance compared with conventional software engineering techniques. Besides, it is difficult to precisely evaluate the degree of flexibility for reuse and evolution of systems provided by the adoption of these SoC techniques. Other serious issues come to mind, such as: is the use of these techniques double-edged? Can these systems suffer a ripple effect, whereby a small change in some part has unexpected and potentially dangerous effects on the whole? The Programming for Separation of Concerns (PSC) track at the 2012 Symposium on Applied Computing (SAC) aims to bring together researchers to share experiences in using SoC techniques, and explore the practical problems of existing tools, environments, etc. The track will address questions like: Can performance degradation be limited? Are unexpected changes dealt with by reflective or aspect-oriented systems? Is there any experience of long term evolution that shows a higher degree of flexibility of systems developed with such techniques? How such techniques cope with architectural erosion? Are these techniques helpful to deal with evolution of legacy systems? Topics ------ Authors are invited to submit original papers. Submissions are encouraged, but not limited, to the following topics: - Software architectures - Configuration management systems - Software reuse and evolution - Performance issues for metalevel and aspect oriented systems - Software engineering tools - Consistency, integrity and security - Generative approaches - Experiences in using reflection, composition filters, aspect- and subject- orientation - Evolution of legacy systems - Reflective and aspect oriented middleware for distributed systems - Modelling of SoC techniques to allow predictable outcomes from their use - Formal methods for metalevel systems Paper Submission ---------------- Original papers from the above mentioned or other related areas will be considered. Only full papers about original and unpublished research are sought. Parallel submission to other conferences or tracks is not acceptable. Papers can be submitted in electronic format via the SAC website (www.softconf.com/c/sac2012/) within 31 August 2011. Please make sure that the authors name and affiliation do not appear on the submitted paper. Peer groups with expertise in the track focus area will blindly review submissions to the track. At least one author of the accepted paper should register and participate in the PSC track. Accepted papers will be published in the annual conference ACM proceedings. The camera-ready version of the accepted paper should be prepared using the ACM format (guidelines will be given on the SAC website). The maximum number of pages allowed for the final papers is six (6), with the option, at additional cost, to add two (2) more pages. A set of papers submitted to the PSC track and not accepted as full papers will be selected as poster papers and published in the ACM proceedings as 2-page papers, with the option, at additional cost, to add one (1) more page. Important Dates --------------- Paper Due August 31, 2011 Author Notification Oct. 12, 2011 Camera Ready Nov. 2, 2011 Please check the web site for updates: www.dmi.unict.it/~tramonta/sac/ -- Resistance is futile. You will be jazzimilated. Scientific site: http://www.lrde.epita.fr/~didier Music (Jazz) site: http://www.didierverna.com EPITA/LRDE, 14-16 rue Voltaire, 94276 Le Kremlin-Bicêtre, France Tel. +33 (0)1 44 08 01 85 Fax. +33 (0)1 53 14 59 22

1 0

New publication: GRETSI 2011: Une approche générique du logiciel pour le traitement d'images préservant les performances
by Roland Levillain 16 May '11

16 May '11

We are happy to announce that the following paper has been accepted for publication at the 23rd Symposium on Signal and Image Processing (GRETSI) that will take place in Bordeaux, France on September 5 - 8, 2011. Nous avons le plaisir de vous annoncer que l'article suivant a été accepté pour publication au 23e colloque GRETSI (Traitement du Signal et des Images), qui aura lieu à Bordeaux en France, du 5 au 8 septembre 2011. Roland Levillain (1,2), Thierry Géraud (1,2) and Laurent Najman (2) Une approche générique du logiciel pour le traitement d'images préservant les performances http://publis.lrde.epita.fr/201109-GRETSI (1) EPITA Research and Development Laboratory (LRDE) (2) Université Paris-Est, Laboratoire d'Informatique Gaspard-Monge, Equipe A3SI, ESIEE Paris De plus en plus d'outils logiciels modernes pour le traitement d'images sont conçus en prenant en compte le problème de la généricité du code, c'est-à-dire la possibilité d'écrire des algorithmes réutilisables, compatibles avec de nombreux types d'entrées. Cependant, ce choix de conception se fait souvent au détriment des performances du code exécuté. Du fait de la grande variété des types d'images existants et de la nécessité d'avoir des implémentations rapides, généricité et performance apparaissent comme des qualités essentielles du logiciel en traitement d'images. Cet article présente une approche préservant les performances dans un framework logiciel générique tirant parti des caractéristiques des types de données utilisés. Grâce à celles-ci, il est possible d'écrire des variantes d'algorithmes génériques offrant un compromis entre généricité et performance. Ces alternatives sont capables de préserver une partie des aspects génériques d'origine tout en apportant des gains substantiels à l'exécution. D'après nos essais, ces optimisations génériques fournissent des performances supportant la comparaison avec du code dédié, allant parfois même jusqu'à surpasser des routines optimisées manuellement. -- Roland Levillain - LRDE/EPITA - A3SI/ESIEE/LIGM/UMLV-Paris Est Laboratoire de Recherche et de Développement de l'EPITA (LRDE) 14-16, rue Voltaire - 94276 Le Kremlin-Bicêtre Cedex Tél. : 01 53 14 59 45 - Fax : 01 53 14 59 13 - www.lrde.epita.fr

1 0

[Seminaire-LRDE] Séminaire Performance et Généricité du LRDE le Mercredi 18 mai 2011
by Jonathan Fabrizio 12 May '11

12 May '11

Chers collègues, La prochaine session du séminaire Performance et Généricité du LRDE (Laboratoire de Recherche et Développement de l'EPITA) aura lieu le Mercredi 18 mai 2011 (14h-16h). Au programme: * 14h: Utilisation des distances tangentes pour la compensation de mouvement : Application au codec Theora -- Jonathan Fabrizio Pour encoder de manière efficace une séquence vidéo, la redondance temporelle est souvent utilisée. Pour cela, le mouvement entre l'image considérée et une image de référence est estimé. Cela permet de générer une prédiction à partir de l'image de référence et seule la différence entre la prédiction et l'image réelle est enregistrée. Pour estimer ce mouvement, les codecs se contentent souvent de suivre l'évolution spatiale de blocs dans l'image. Ils associent, pour chaque bloc de l'image considérée, un bloc similaire dans un voisinage proche dans l'image de référence. Nous présentons ici une méthode originale pour réaliser cette prédiction par compensation de mouvement. Notre méthode utilise les distances tangentes. Cela permet non seulement d'estimer l'évolution de la position des blocs de l'image mais en partie aussi l'évolution du bloc lui-même. Nos prédictions sont donc de meilleure qualité. Utilisée dans l'encodage de séquences, on peut espérer un gain de compression non négligeable. Pour valider l'efficacité de la méthode, nous intégrons cette méthode dans le codec Theora et mesurons son efficacité en comparant les résultats obtenus avec notre méthode et ceux obtenus par une stratégie classique (le block-matching). -- Jonathan Fabrizio est enseignant-chercheur au LRDE depuis l'été 2009. Il travaille actuellement sur la localisation et l'extraction automatique de texte dans les images. Pour plus de renseignements, consultez http://seminaire.lrde.epita.fr/. L'entrée du séminaire est libre. Merci de bien vouloir diffuser cette information le plus largement possible. _______________________________________________ Seminaire mailing list Seminaire(a)lrde.epita.fr https://www.lrde.epita.fr/mailman/listinfo/seminaire

1 0

CDR #7: Clarification Proposal for CLHS 22.3
by Didier Verna 05 Apr '11

05 Apr '11

Hello, I'm pleased to announce that my "Clarification Proposal for CLHS 22.3" has been accepted, finalized and published at the Common Document Repository. It can now be referred to as CDR #7. The abstract is given below: Section 22.3 "Formatted Output" of the Common Lisp Hyperspec describes the syntax and semantics of format directives. We believe that the standard is underspecified. We propose to clarify that section, and suggest that Common Lisp implementations conform to the current behavior of CMU-CL, CCL, CLISP, Allegro and LispWorks. -- Resistance is futile. You will be jazzimilated. Scientific site: http://www.lrde.epita.fr/~didier Music (Jazz) site: http://www.didierverna.com EPITA/LRDE, 14-16 rue Voltaire, 94276 Le Kremlin-Bicêtre, France Tel. +33 (0)1 44 08 01 85 Fax. +33 (0)1 53 14 59 22

1 0

[CfP] DLS 2011 - 7th Dynamic Languages Symposium
by Didier Verna 05 Apr '11

05 Apr '11

Dynamic Languages Symposium 2011 Co-located with SPLASH 2011 In association with ACM SIGPLAN Portland, Oregon, USA, October 24, 2011 http://www.dynamic-languages-symposium.org/dls-11/ *** Call for papers *** The 7th Dynamic Languages Symposium (DLS) at SPLASH 2011 is a forum for discussion of dynamic languages, their implementation and application. While mature dynamic languages including Smalltalk, Lisp, Scheme, Self, Prolog, and APL continue to grow and inspire new converts, a new generation of dynamic scripting languages such as Python, Ruby, PHP, Tcl, Lua, and JavaScript are successful in a wide range of applications. DLS provides a place for researchers and practitioners to come together and share their knowledge, experience, and ideas for future research and development. DLS 2011 invites high quality papers reporting original research, innovative contributions or experience related to dynamic languages, their implementation and application. Accepted Papers will be published in the ACM Digital Library. Areas of interest include but are not limited to: - Innovative language features and implementation techniques - Development and platform support, tools - Interesting applications - Domain-oriented programming - Very late binding, dynamic composition, and runtime adaptation - Reflection and meta-programming - Software evolution - Language symbiosis and multi-paradigm languages - Dynamic optimization - Hardware support - Experience reports and case studies - Educational approaches and perspectives - Object-oriented, aspect-oriented, and context-oriented programming * Submissions and proceedings * We invite original contributions that neither have been published previously nor are under review by other refereed events or publications. Research papers should describe work that advances the current state of the art. Experience papers should be of broad interest and should describe insights gained from substantive practical applications. The program committee will evaluate each contributed paper based on its relevance, significance, clarity, and originality. Accepted papers will be published in the ACM Digital Library. Papers are to be submitted electronically at http://www.easychair.org/conferences?conf=dls2011 in PDF format. Submissions must not exceed 12 pages and need to use the ACM format, templates for which can be found at http://www.acm.org/sigs/pubs/proceed/template.html. * Important dates * Submission of papers: June 17, 2011 (hard deadline) Author notification: July 19, 2011 Final versions due: August 19, 2011 DLS 2011: October 24, 2011 SPLASH 2011: October 22-27, 2011 * Program chair * Theo D'Hondt, Software Languages Lab, Vrije Universiteit Brussel, Belgium * Program committee * Andrew Black, Portland State University, USA William R. Cook, University of Texas at Austin, USA Marc Feeley, University of Montreal, Canada Roberto Ierusalimschy, PUC-Rio, Brazil Michele Lanza, University of Lugano, Switzerland Hidehiko Masuhara, University of Tokyo, Japan Mira Mezini, University of Darmstadt, Germany Mark Miller, Google, USA Manuel Serrano, INRIA Nice, France Laurence Tratt, Middlesex University, UK David Ungar, IBM, USA Didier Verna, EPITA Research and Development Laboratory, France -- Resistance is futile. You will be jazzimilated. Scientific site: http://www.lrde.epita.fr/~didier Music (Jazz) site: http://www.didierverna.com EPITA/LRDE, 14-16 rue Voltaire, 94276 Le Kremlin-Bicêtre, France Tel. +33 (0)1 44 08 01 85 Fax. +33 (0)1 53 14 59 22

1 0

[Final CfPart] European Lisp Symposium, Hamburg, March 31 - April 1
by Didier Verna 09 Mar '11

09 Mar '11

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4th European Lisp Symposium Special Focus on Parallelism & Efficiency March 31 - April 1st, 2011 TUHH, Hamburg University of Technology Hamburg, Germany http://www.european-lisp-symposium.org/ Sponsors: EPITA, TUHH, Lispworks, Franz Inc., NovaSparks and Freiheit ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ News: ~~~~~ * The final program is now online. * The early registration deadline is in 3 days, so hurry! Registration will still be possible afterwards. Invited Speakers: ~~~~~~~~~~~~~~~~~ Craig Zilles -- Compiling for the common case Marc Battyani -- Reconfigurable computing on steroids Apostolos Syropoulos -- Scala: an OO surprise Scope ~~~~~~ The purpose of the European Lisp Symposium is to provide a forum for the discussion and dissemination of all aspects of design, implementation and application of any of the Lisp dialects, including Common Lisp, Scheme, Emacs Lisp, AutoLisp, ISLISP, Dylan, Clojure, ACL2, ECMAScript, Racket, SKILL and so on. We encourage everyone interested in Lisp to participate. The European Lisp Symposium 2011 invites high quality papers about novel research results, insights and lessons learned from practical applications, and educational perspectives. We also encourage submissions about known ideas as long as they are presented in a new setting and/or in a highly elegant way. This year's focus will be directed towards "Parallelism & Efficiency". Programme Chair ~~~~~~~~~~~~~~~~ Didier Verna - EPITA Research and Development Laboratory, France Local Chair ~~~~~~~~~~~~ Ralf Moeller - Hamburg University of Technology, Germany Programme Committee ~~~~~~~~~~~~~~~~~~~~ Antonio Leitao - Instituto Superior Tecnico/INESC-ID, Portugal Christophe Rhodes - Goldsmiths College, University of London, UK David Edgar Liebke - Relevance Inc., USA Didier Verna - EPITA Research and Development Laboratory, France Henry Lieberman - MIT Media Laboratory, USA Jay McCarthy - Brigham Young University, USA Jose Luis Ruiz Reina - Universidad de Sevilla, Spain Marco Antoniotti - Universita Milano Bicocca, Italy Manuel Serrano - INRIA, France Michael Sperber - DeinProgramm, Germany Pascal Costanza - Vrije Universiteit of Brussel, Belgium Scott McKay - ITA Software, USA -- Resistance is futile. You will be jazzimilated. Scientific site: http://www.lrde.epita.fr/~didier Music (Jazz) site: http://www.didierverna.com EPITA/LRDE, 14-16 rue Voltaire, 94276 Le Kremlin-Bicêtre, France Tel. +33 (0)1 44 08 01 85 Fax. +33 (0)1 53 14 59 22

1 0