Vous êtes ici : Réunions » Réunion

Identification

Identifiant: 
Mot de passe : 

Mot de passe oublié ?
Détails d'identification oubliés ?

Compression et sécurisation pour les applications immersives

Nous vous rappelons que, afin de garantir l'accès de tous les inscrits aux salles de réunion, l'inscription aux réunions est gratuite mais obligatoire.

Inscriptions closes à cette réunion.

Inscriptions

21 personnes membres du GdR ISIS, et 13 personnes non membres du GdR, sont inscrits à cette réunion.
Capacité de la salle : 100 personnes.

Annonce

Les technologies immersives apportent de nouvelles façons d'utiliser les contenus visuels dans nos applications de la vie quotidienne. Ils jouent un rôle important dans l'amélioration des expériences d'apprentissage, en favorisant la participation, la collaboration, la créativité et l'engagement. Les applications sont diverses incluant la 3D stéréoscopique, la réalité virtuelle (VR), la réalité augmentée (AR), la réalité mixte (MR), etc. Les nouveaux types de contenus comme les nuages de points, les champs de lumière (lightfields), les images 360-deg, contribuent significativement dans l'expansion de ce domaine et à fournir une expérience de plus en plus confortable à l'utilisateur.

Cependant, comme dans toute chaîne d'imagerie, ces nouveaux contenus et ces nouvelles applications font émerger de nouveaux besoins notamment en termes de compression et de sécurisation des contenus. La généralisation de l'utilisation de l'apprentissage profond a également contribué à intensifier les travaux de recherche dans ces domaines.

Cette journée du GDR-ISIS organisée par le thème D vise à créer un espace de partage scientifique sur les avancées récentes de ce domaine entre les chercheurs seniors, juniors et les industriels. Elle sera scindée en deux parties : la première orientée sécurisation en lien avec le GdR Sécurité et la seconde sur les problématiques de compression.

Organisateurs

Programme

9h45-12h00 : Compression de contenus immersifs

11h50-12h30 : Qualité des contenus immersifs

14h00-15h25 : Sécurisation des données 3D

15h45-16h45: Sécurisation de documents imprimés, d'images, et de vidéos

Pour participer à la partie sécurité vous devrez vous inscrire également sur le site du gdr sécurité : https://gdr-secu-jn2021.sciencesconf.org/

3D watermarking - from graphical models to 3D printing

Adrian BORS, University of York, UK

Abstract : 3D watermarking or steganography refers to embedding information into 3D objects, such as graphical models. The information associated with the 3D watermark is encoded through perturbations in the content of the object, which cannot be observed, and we should be able to retrieve the code associated with such perturbations. Moreover, the watermark should be robust to various transformations and processing that the object may undergo, or to specific attacks intent to destroy the embedded code. In the talk we will discuss the constrains that a 3D watermark must fulfil and the challenges to retrieve it safely. 3D watermarking can be used for copyright protection, but also for embedding useful information associated with the object. Fragile watermarks can be also used for the authentication of the object. After discussing some watermarking methods for graphical models we will discuss about 3D steganalysis, which represents the method to identify whether watermarks have been embedded into 3D objects by performing a series of tests to the graphical object. We will discuss how 3D watermarking can be applied in real objects through 3D printing and its challenges.

Short Biography : Adrian G. Bors received the MSc degree in Electronics Engineering from the Polytechnic Univ. of Bucharest, Romania, in 1992, and the Ph.D. degree in Informatics from the Univ. of Thessaloniki, Greece in 1999. In 1999 he joined the Department of Computer Science, Univ. of York, U.K., where he is currently a lecturer. Dr. Bors held temporary positions at Tampere Univ. of Technology, Finland, Univ. of California at San Diego (UCSD), and at the Univ. of Montpellier, France. Dr. Bors has authored and co-authored more than 130 research papers including 30 in journals. His research interests include computer vision, computational intelligence and image processing. Dr. Bors was an associate editor of IEEE Trans. Image Processing (2010 ? 2014) and of IEEE Trans. Neural Networks (2001 ? 2009). He was a co-guest editor for a special issue on Machine Vision for the International Journal for Computer Vision in 2018, and for the Journal of Pattern Recognition in 2015. Dr. Bors was a member of the organisation committees for WIFS 2021, IPTA 2020, ICIP 2018, BMVC 2016, IPTA 2014, CAIP 2013 and ICIP 2001. He is a Senior Member of the IEEE.

Overview of technologies for immersive visual experiences

Prof. Marek Domanski, Poznan University of Technology, Institute of Multimedia Telecommunications

Abstract : The lecture is focused on the state-of-the-art techniques for immersive video. The lecture is not intended for explanation of mathematical details, but the lecture is aimed at general description of the problems, recent research results, existing solutions, technical limitations, examples of practical solutions, standardization, and expected future development.
The lecture content:

  1. Immersive video ? what it is.
  2. Overview of virtual navigation / free-viewpoint video / free-viewpoint television video processing path : Acquisition, calibration/ system parameter estimation, preprocessing (model estimation, depth estimation)/ transmission/ edge server/ user terminal. Example of a system in a sports hall.
  3. System architecture variants and compression in the individual links. Requirements.
  4. Compression solutions available : Technology available, compression standards (HEVC, 3D-HEVC,?) , new Immersive Video Coding standard (ISO MPEG-I Part 12).
  5. View synthesis, rendering and presentation.

Biography : MAREK DOMA?SKI received M.Sc., Ph.D. and Habilitation degrees from Pozna? University of Technology, Poland in 1978, 1983 and 1990, respectively. Since 1993, he is a professor at Pozna? Univ. of Technology, where he is the director of Institute of Multimedia Telecommunications. He coauthored one of the very first AVC decoders for tv set-top boxes (2004) as well as highly ranked technology proposals to MPEG for scalable video compression (2004), 3D video coding (2011), and immersive video coding (2019). He authored 3 books and over 300 papers in journals and conference proceedings. 16 patents granted by European Patent Office and United States Patent and Trademark Office with him as a co-inventor. He is active in international standardization expert group MPEG, currently by MPEG-I (MPEG Immersive). He serves as a co-editor for MPEG Test Model for Immersive Video. He promoted (directed) 23 candidates to Doctor degree. The contributions were mostly on image, video and audio compression, virtual navigation, free-viewpoint television, image processing, multimedia systems, 3D video and color image technology, digital filters and multidimensional signal processing. He was leading many research projects funded by Polish institutions (science foundations and agencies), EU and NATO, as well as industry and industrial institutes from Poland, China, Japan, Rep. of Korea, Germany, USA etc. He is He was General Chairman/Co-Chairman and host of several international conferences: Picture Coding Symposium, PCS 2012; IEEE Int. Conf. Advanced & Signal based Surveillance, AVSS 2013, European Signal Processing Conference, EUSIPCO 2007; 73rd and 112nd Meetings of MPEG; Int. Workshop on Signals, Systems and Image Processing, IWSSIP 1997 and 2004; Int. Conf. Signals and Electronic Systems, ICSES 2004.

Les exemples adversaires, la partie visible de l'iceberg Sécurité de l'apprentissage automatique?

Teddy Furon, IRISA, Rennes.

Abstract Biography

Résumés des contributions

Date : 2021-07-01

Lieu : En visio


Thèmes scientifiques :
D - Télécommunications : compression, protection, transmission

Inscriptions closes à cette réunion.

(c) GdR 720 ISIS - CNRS - 2011-2022.