Cette page appartient aux archives web de l'EPFL et n'est plus tenue à jour.
This page belongs to EPFL's web archive and is no longer updated.

News of the MMSPG

New PhD granted to David Marimon
David Marimon successfully defends his PhD thesis entitled
Advances in Top-Down and Bottom-Up Approaches to Video-Based Camera Tracking

The public defense will take place on Thursday 6th Dec in room ELA1 (http://plan.epfl.ch/?room=ELA1)

Abstract of his thesis
Video-based camera tracking consists in trailing the three dimensional pose followed by a mobile camera using video as sole input. In order to estimate the pose of a camera with respect to a real scene, one or more three dimensional references are needed. By comparing what is seen by a camera with what is geometrically known from reality, it is possible to recover the pose of the camera that is sensing these references.

In this thesis, we investigate the problem of camera tracking at two levels. Firstly, we work at the low level of feature point recognition. Feature points are used as references for tracking and we propose a method to robustly recognise them. More specifically, we introduce a rotation-discriminative region descriptor and an efficient matching method. Secondly, we have worked at the higher level of camera tracking and propose a fusion of top-down (TDA) and bottom-up approaches (BUA). We combine marker-based tracking using a BUA and feature points recognised from a TDA into a particle filter. The goal of the fusion is to take advantage of their compensated strengths. In particular, we are interested in covering automatic initialisation, automatic recovery after loss of track, and tracking beyond references known a priori.

Experiments have been performed at the two levels of investigation. Firstly, tests have been conducted to evaluate the performance of the recognition method proposed. A comparison with similar techniques in the state of the art depicts the equal or even higher precision of our method with much lower computational cost. Secondly, experimental assessment of the tracking system is also conducted. Results show that the fusion tracker solves characteristic failure modes of the two combined approaches. A comparison with similar trackers shows competitive accuracy. In addition, all the three capabilities stated earlier are fulfilled only in our tracker.

The camera tracking system has a potential application in the robotics domain. It has been successfully used as a man-machine interface, applied in Augmented Reality environments and tested in public events. In particular, the system has been used by students of the Ecole Cantonale d'Art de Lausanne (ECAL) with the purpose of conceiving new interaction concepts. Moreover, in collaboration with ECAL and fabric|ch (studio for architecture & research), we have jointly developed the Augmented interactive Reality Toolkit (AiRToolkit).
Posted by David Marimon Sanjuan at 10:09
Sixth International Workshop on Content-Based Multimedia Indexing, CBMI 2008, London, United Kingdom, June 18-10 June, 2008
Sixth International Workshop on Content-Based Multimedia Indexing, CBMI 2008, London, United Kingdom, June 18-10 June, 2008
http://cbmi08.qmul.net/
Posted by David Marimon Sanjuan at 10:03
Prof. Touradj Ebrahimi gives a seminar on JPSearch at the Hong Kong Baptist University.
Prof. Touradj Ebrahimi gives a seminar on JPSearch at the Hong Kong Baptist University.
http://www.comp.hkbu.edu.hk/~rpoon/final/seminar/seminar/2007/251007/20071025.html
Posted by David Marimon Sanjuan at 10:02
IEEE International Conference on Multimedia & Expo, June 23-26, 2008, Hannover, Germany
IEEE International Conference on Multimedia & Expo, June 23-26, 2008, Hannover, Germany
http://www.icme2008.org/
Posted by David Marimon Sanjuan at 10:00
Picture Coding Symposium, 7-9 Nov. 2007, Lisbon, Portugal
Picture Coding Symposium, 7-9 Nov. 2007, Lisbon, Portugal
http://www.img.lx.it.pt/pcs2007/
Posted by David Marimon Sanjuan at 9:59
New PhD students joined the team: Peter Vajda and Ashkan Yazdani
New PhD students joined the team: Peter Vajda and Ashkan Yazdani
Posted by David Marimon Sanjuan at 9:55
Special Issue on Visual Media Quality Assessment at the IEEE Journal of Selected Topics in Signal Processing.
Call for Papers


IEEE Journal of Selected Topics in Signal Processing
IEEE Signal Processing Society
Special Issue on Visual Media Quality Assessment


With the increasing demand for digital image and video technologies in applications as broad as entertainment and communications, security, monitoring, and medical imaging, there is a growing need for the automatic assessment of the quality of visual media. Many factors can affect and impair the quality of visual media including compression, transmission, protection, display, printing, acquisition and reproduction systems. Automatic visual media quality assessment is crucial for monitoring and controlling the visual quality in existing and emerging multimedia systems, and has the potential to impact next-generation systems by providing objective metrics for use during the design and testing stages and by reducing the need for extensive evaluation with human subjects.

Visual media quality assessment aims at quantifying the quality of visual media, including still pictures, image sequences, 3D visual data, and 3D models, by means of quality metrics. These metrics vary with the considered applications, and range from metrics that measure specific visual impairments to those that assess the overall visual quality in the presence of various impairments. For applications and products that target human consumers, it is desirable to have metrics that will predict the perceived visual quality as measured with human subjects. Visual quality assessment metrics can be further divided into full-reference, reduced-reference, and no-reference quality metrics. Full-reference visual quality metrics compare the to-be-assessed visual media to a reference, which is typically the original visual data. In many applications where the original visual data is not available, reduced-reference and no-reference metrics are used. Reduced-reference metrics make use of a set of reference features or characteristics, which could have been extracted from the original visual data. No-reference quality metrics attempt to predict the visual quality without any reference, which is very useful in practice but very challenging.

A great deal of interest and research have been devoted to the design and development of visual quality metrics, particularly full-reference and reduced-reference metrics for image quality assessment. However, for many applications, reliable automatic visual quality assessment is lacking, particularly those requiring no-reference visual quality assessment. In addition, there is a need for methods that can reliably assess the visual quality of video and other 3D visual media. The motivation for this special issue is to highlight the importance, challenges, and applications of visual media quality assessment and its interdisciplinary nature which includes vision science, optics, color science, signal processing, psychology, and biology. Our goal is to feature recent advances in the area of automatic visual media quality assessment, including theoretical, experimental, and computational methods and results. We invite researchers to submit original papers describing new approaches in all areas related to automatic visual media quality assessment including, but not limited to, the following topics:
- Global and impairment-specific visual quality assessment metrics
- Full-reference, reduced-reference, and no-reference visual quality assessment of still-pictures and video.
- Visual quality assessment of 3D visual data and 3D models
- Visual quality assessment of High-Definition image and video content
- Statistical methods for automatic visual quality assessment
- Perceptually/Biologically-inspired automatic visual quality assessment
- Visual quality metrics for specific applications

Submission procedure
Prospective authors can find submission information at http://www.ece.byu.edu/jstsp. Submitted manuscripts should not have been previously published nor be currently under consideration for publication elsewhere. Authors are advised to follow the Author's Guide for the formats of manuscripts submitted to the IEEE Transactions on Signal Processing as detailed at http://ewh.ieee.org/soc/sps/tsp/. The manuscript will undergo a standard peer review process.
Manuscript submissions due: April 30, 2008
First review completed: July 31, 2008
Revised manuscripts due: September 15, 2008
Second review completed: October 31, 2008
Final manuscript due: November 30, 2008

Lead Guest Editor
Lina Karam, Arizona State University, Tempe, Arizona (karam@asu.edu)

Guest Editors
Touradj Ebrahimi, EPFL, Lausanne, Switzerland (touradj.ebrahimi@epfl.ch)
Sheila Hemami, Cornell University, Ithaca, New York (hemami@ece.cornell.edu)
Thrasos Pappas, Northwestern University, Evanston, Illinois (pappas@ece.northwestern.edu)
Robert Safranek, Benevue, Warren, New Jersey (rjs@@benevue.com)
Zhou Wang, University of Waterloo, Waterloo, Ontario, Canada (z.wang@ece.uwaterloo.ca)
Andrew B. Watson, NASA Ames Research Center, Moffett Field, California (andrew.b.watson@nasa.gov)
Posted by David Marimon Sanjuan at 9:44
COST 292 Special Session: Towards Automated Robust Vision-based Surveillance
Call for Papers
COST 292 Special Session
Towards Automated Robust Vision-based Surveillance


Organisers: Dr. Noel E. O'Connor (DCU), Dr. Marzia Corvaglia (Uni. Brescia)

The members of the COST 292 action (http://www.cost292.org/) have a long tradition of involvement in WIAMIS, dating back to the establishment of this event. Since 2005, the action has been invited to organize a special session and this tradition will be continued for WIAMIS 2008. This year, the special session will focus on the newly identified priority areas of this action that fall into the broad area of automated visual surveillance. This new focus has been prioritized on the basis of the observation that research in Europe in this area is fragmented with no common forum for researchers working on different aspects of this broad topic. The goal of this special session is to help bring together researchers working on this topic and to help stimulate integration in this area.


Programme Subcommittee
Selim Aksoy, Bilkent Uni., Turkey
Michael Ansorge, EPFL, Switzerland
Yannis Avrithis, NTUA, Greece
Jenny Benois-Pineau, LABRI, France
Licia Capodiferro, FUB, Italy
Touradj Ebrahimi, EPFL, Switzerland
Alan Hanjalic, TU Delft, NL
Ebroul Izquierdo, QMUL, U.K.
Yiannis Kompatsiaris, ITI, Greece
Riccardo Leonardi, Uni. Brescia, Italy
Gorka Marcos Ortego, VICOMTech, Spain
Branimir Reljin, Uni. Belgrade, Serbia
Thomas Sikora, TU Berlin, Germany
Walter Stechele, TU Munchen, Germany


Topics of interest include, but are not limited to:
- Visual computation: background subtraction, object detection and tracking, body and face detection, face recognition,
- Event and activity modeling and analysis: human action modeling and recognition, crowd behaviour analysis;
- Multiple stream analysis: fusion across similar and/or differing visual modalities, augmenting visual analysis with other modalities such as audio;
- Architectures: video surveillance network architectures, new sensors and cameras;
- Annotation, Indexing and Storage: archival & retrieval of sensor data and extracted semantics, query paradigms and languages;
- Coding: event-/object-based compression for surveillance, scalable coding, distributed coding;
- Applications: crime prevention, aerial imaging (satellite, infrared), public transport systems, privacy protection, anonymity enforcement;


Important Dates
Paper Submission: Dec. 14, 2007
Review Results: Jan. 11, 2008
Camera-ready Papers: Feb. 01, 2008
Papers should conform to the usual WIAMIS guidelines for authors. All papers will be peer reviewed by the WIAMIS Technical Programme Committee and a special Programme Subcommittee set up specifically for this special session.


For more information contact:
Dr. Noel E. O'Connor, oconnorn (at) eeng (dot) dcu (dot) ie
Dr. Marzia Corvaglia, marzia (dot) corvaglia (at) ing (dot) unibs (dot) it
Posted by David Marimon Sanjuan at 9:42
Prof. Touradj Ebrahimi, plenary speaker at the IEEE 2007 International Workshop on Multimedia Signal Processing, 1-3 Oct.2007, Chania, Greece
Prof. Touradj Ebrahimi, plenary speaker at the IEEE 2007 International Workshop on Multimedia Signal Processing, 1-3 Oct.2007, Chania, Greece
http://www.mmsp2007.org/
Posted by David Marimon Sanjuan at 9:40
Discover workshop: Recent Advances in Distributed Video Coding, 6 Nov. 2007, Lisbon, Portugal
Discover workshop: Recent Advances in Distributed Video Coding, 6 Nov. 2007, Lisbon, Portugal
http://www.discoverdvc.org/Workshop_Programme.html
Posted by David Marimon Sanjuan at 9:39
Page : « Previous 1 2 3 4 5 6 7 8 9 10 11 Next »