Archives pour la catégorie publication

publication

Visual Content Indexing and Retrieval with Psycho-Visual Models

I’m very pround to announce my chapter  » Information – theoretical model for saliency prediction. Application to Attentive CBIR   » in

Visual Content Indexing and Retrieval with Psycho-Visual Models

Edited by  Jenny Benois-Pineau , Edited by  Patrick Le Callet
Abstract : « This work presents an original informational approach to extract visual information, model attention and evaluate the efficiency of the results.
Even if the extraction of salient and useful information, i.e. observation, is an elementary task for human and animals, its simulation is still an open problem in computer vision. In this article, we define a process to derive specific and optimal laws to extract visual information and by the way model information without any constraints or a priori. Starting from saliency definition and measure through the prism of information theory, we present a framework in which we develop an ecological inspired approach to model visual information extraction. We theoretically demonstrate some results previously presented, for instance, in spites of being fast and highly configurable, our model is as plausible as existing models designed for high biological fidelity. It proposes an adjustable trade-off between nondeterministic attentional behavior and properties of stability, reproducibility and reactiveness. We apply this approach to enhance the performance in an object recognition task.
As a conclusion, this article proposes a theoritical framework to derive an optimal model validated by many experimentations. »
publication

« From Human Attention to Computational Attention – A Multidisciplinary Approach »

Dear all,

I hope you all had great holidays and are ready to get into a new year of work 🙂
Our book « From Human Attention to Computational Attention – A Multidisciplinary Approach » was published at Springer Series in Cognitive and Neural Systems. You can find it here: http://www.springer.com/la/book/9781493934331

This both accessible and exhaustive book will help to improve modeling of attention and to inspire innovations in industry. It introduces the study of attention and focuses on attention modeling, addressing such themes as saliency models, signal detection and different types of signals, as well as real-life applications. The book is truly multi-disciplinary, collating work from psychology, neuroscience, engineering and computer science, amongst other disciplines.

What is attention? We all pay attention every single moment of our lives. Attention is how the brain selects and prioritizes information. The study of attention has become incredibly complex and divided: this timely volume assists the reader by drawing together work on the computational aspects of attention from across the disciplines. Those working in the field as engineers will benefit from this book’s introduction to the psychological and biological approaches to attention, and neuroscientists can learn about engineering work on attention. The work features practical reviews and chapters that are quick and easy to read, as well as chapters which present deeper, more complex knowledge. Everyone whose work relates to human perception, to image, audio and video processing will find something of value

in this book, from students to researchers and those in industry.

publication

Accepted paper in JCDL conf (rank A*)

I am pleased to inform that our submission to JCDL 2016 has been accepted. http://www.jcdl2016.org/

TITLE: User Activity Characterization in a Cultural Heritage Digital Library System
AUTHORS: Cyrille Suire, Axel Jean-Caurant, Vincent Courboulay, Jean-Christophe Burie and Pascal Estraillier

The ACM/IEEE-CS Joint Conference on Digital Libraries (JCDL) is a major international forum focusing on digital libraries and associated technical, practical, and social issues. Participation is sought from all parts of the world and from the full range of disciplines and professions involved in digital library research and practice, including computer science, information science, librarianship, archival science and practice, museum studies and practice, technology, medicine, social sciences, and humanities.

Non classé, publication

[GdR ISIS] Saillance visuelle et applications au tatouage et à la compression d’images et de vidéos

Réunion du GdR ISIS

  • Titre : Saillance visuelle et applications au tatouage et à la compression d?images et de vidéos
  • Dates : 2016-05-31
  • Lieu : Télécom Paristech – Amphi Opale

Annonce :

Saillance visuelle et applications au tatouage et à la compression d’images et de vidéos

Journée commune Thèmes B et D

Ces dernières années, de nombreuses études ont porté sur l’attention visuelle débouchant sur un nombre important de nouvelles méthodes d’extraction de la saillance visuelle ainsi que son utilisation dans diverses applications aussi bien comme outil de guidage (extraction d’objets, suivi, …) ou bien son utilisation comme étape intermédiaire permettant de focaliser uniquement sur certaines régions (qualité d’image, tatouage, …). Ainsi, à travers ces différentes études, la pertinence d’utiliser ces modèles a été largement mise en évidence.

La thématique principale que nous souhaitons développer est donc celle de la saillance visuelle et de son exploitation. Parmi les thématiques émergentes exploitant la saillance, on citera en particulier les méthodes de tatouage numérique d’images ou de vidéos. Dans le cadre de cette journée, entre thème B et thème D, qui fait suite aux deux journées précédentes sur la saillance visuelle, une attention particulière sera donc apportée, cette année, aux applications industrielles ou académiques de type sécurité ou compression des images, sans pour autant exclure les autres domaines.

Orateurs invités :

  • Frédéric Dufaux, Directeur de recherche CNRS, LTCI, Télécom Paristech
  • Vincent Courboulay, Maitre de conférence-HDR, L3i, Université de La Larochelle

Programme préliminaire :

En cours d’élaboration.

Organisateurs :

  • Aladine Chetouani, PRISME, Université d’Orléans
  • Vincent Courboulay, L3i, Université de La Larochelle
  • Christine Fernandez, XLIM, Université de Poitiers
  • William Puech, LIRMM, Montpellier

Appel à contributions :

Les personnes souhaitant présenter leurs travaux sont priées d’envoyer un résumé d’une page maximum (format pdf) avant le 6 mai 2016 aux organisateurs:

Lien : http://gdr-isis.fr/index.php?page=reunion&idreunion=310

publication

new article in Creative technologies

Improvement of natural image search engines results by emotional filtering

Abstract
With the Internet 2.0 era, managing user emotions is a problem that more and more actors are interested in. Historically, the first notions of emotion sharing were expressed and defined with emoticons. They allowed users to show their emotional status to others in an impersonal and emotionless digital world. Now, in the Internet of social media, every day users share lots of content with each other on Facebook, Twitter, Google+ and so on. Several new popular web sites like FlickR, Picassa, Pinterest, Instagram or DeviantArt are now
specifically based on sharing image content as well as personal emotional status. This kind of information is economically very valuable as it can for instance help commercial companies sell more efficiently. In fact, with this king of emotional information, business can made where companies will better target their customers needs, and/or even sell them more products. Research has been and is still interested in the mining of emotional information from user data  since then. In this paper, we focus on the impact of emotions from images that have been collected from search image engines. More  specifically our proposition is the creation of a filtering layer applied on the results of such image search engines. Our peculiarity relies in the fact that it is the first attempt from our knowledge to filter image search engines results with an emotional filtering approach.

 

Non classé, publication

Présentation invitée au colloque Emotions, mémoire et contenus numériques: une approche interdisciplinaire

Intitulé de mon intervention :

Amélioration des résultats d’un moteur de recherche d’images naturelles par filtrage émotionnel

 

Un colloque organisé par l’IRCCyN et le LPPL lors des Journées Scientifiques de l’Université de Nantes

L’objectif de ce colloque est de faire se rencontrer les communautés STIC et psychologie
afin de débattre et échanger autour d’un sujet commun : les émotions et la mémorisation
dans le cadre de contenus numériques (images, vidéos, dispositifs interactifs, etc.). Cette
rencontre pourra également servir de tremplin afin de renforcer et animer une communauté
interdisciplinaire sur ce sujet tant au niveau local qu’international.
Les contributions provenant des domaines de l’informatique, des sciences cognitives et de
la psychologie dans les thématiques suivantes sont les bienvenues :
• Emotions et contenus numériques;
• Mémoire et contenus numériques;
• Mémoire et émotions.

 

Abstract de mon intervention :

Avec l’Internet 2.0, la gestion des émotions de l’utilisateur est un problème de plus en plus pregnant.

Historiquement, les premières tentatives de partage émotionnel furent les émoticônes. Elles ont permis aux utilisateurs d’afficher leur état émotionnel à d’autres dans un monde numérique impersonnel et froid.

De nos jours, les utilisateurs partagent leur vie sur Facebook, Twitter, Google+ et ainsi de suite… et plusieurs sites web, parmi les plus populaires, sont maintenant spécifiquement basée sur le partage de contenu d’images : Flickr, Picasa,Pinterest, Instagram, DeviantArt par exemple.

Ce contenu comprend implicitement leur statut personnel et émotionnel.

Ce type d’information est économiquement très précieux, il devient un nouvel or noir et peut par exemple aider les entreprises commerciales à vendre plus et mieux.

Dans cet exposé, notre proposition concerne la création d’une couche de filtrage émotionnel appliquée aux résultats de moteurs de recherche d’images. C’est une première tentative de filtrage émotionnel d’images.

 

publication

Accepted for EUSIPCO2015

Our paper « A CBIR based evaluation framework for visual attention models » is accepted for publication in EUSIPCO 2015.

http://www.eusipco2015.org 

abstract :

The computational models of visual attention, originally proposed as cognitive models of human attention, nowadays are being used as front-ends to numerous vision systems like automatic object recognition.  These systems are generally evaluated against eye tracking data, or manually segmented salient objects in images. We previously showed that this comparison can lead to different rankings depending on which of the two ground truths is used. These findings suggest that the saliency models ranking might be different for each application and the use of eye-tracking rankings to choose a model for a given application is not optimal. Therefore, in this paper, we propose a new saliency evaluation framework optimized for object recognition. This paper aims to answer the question: 1) Is the application-driven saliency models rankings consistent with classical ground truth like eye-tracking? 2) If not, which saliency models one should use for the precise CBIR applications?

 

publication

Accepted for icdar 2015

Our paper « Multiresolution Approach Based on Adaptive Superpixels for Administrative Documents Segmentation into Color Layers » is accpeted for ICDAR 2015 http://2015.icdar.org/ 

abstract : Administrative document images are usually pro-
cessed in black and white what generates many problems due
to the errors related to the binarization. Besides all semantic
information provided by the color is lost. Document images have
a rich and highly variable content. The presence of false colors
and artefacts introduced by the scanning and the compression
alter the segmentation of the regions. Problems arise when there
is no correspondence between the point clouds which are detected
in a color space and the real regions of an image. In order to
help the segmentation, we propose the extraction of the main
colors of an image as a set of binary layers. Due to the industrial
context, our approach has to run unsupervised on a generic
dataset of color administrative documents. The originality of this
approach is the use of a multiresolution analysis to detect the
number of colors automatically. At a low resolution, a set of local
regions is obtained thanks to a SLIC-based approach which takes
into account the structure of documents and which combines
both colorimetric information and spatial information. Then, a
merging stage is applied on each resolution separately based on
the colors which have been extracted at a lower resolution. This
contribution can both feed the traditional process and exploit
colorimetric information.

AUTHORS: Elodie Carel, Jean-Christophe Burie, Vincent Courboulay, Jean-Marc Ogier and Vincent Poulain d’Andecy