Article de revue avec comité de lecture (3)
AMMAR Marwa, MITREA Mihai, HASNAOUI Marwen, LE CALLET Patrick
MPEG-4 AVC stream-based saliency detection : application to robust watermarking. Signal processing : image communication, february 2018, vol. 60, pp. 116-130
abstract
By bridging uncompressed-domain saliency detection and MPEG-4 AVC compression principles, the present paper advances a methodological framework for extracting the saliency maps directly from the stream syntax elements. In this respect, inside each GOP, the intensity, color, orientation and motion elementary saliency maps are related to the energy of the luma coefficients, to the energy of chroma coefficients, to the gradient of the prediction modes and to the amplitude of the motion vectors, respectively. The three spatial saliency maps are pooled according to an average formula, while the static-temporal fusion is achieved by six different formulas. The experiments consider both ground-truth and applicative evaluations. The ground-truth benchmarking investigates the relation between the predicted MPEG-4 AVC saliency map and the actual human saliency, captured by eye-tracking devices. It is based on two corpora (representing density fixation maps and saccade locations), two objective criteria (related to the closeness between the predicted and the real saliency maps and to the difference between the behavior of the predicted saliency map in fixation and random locations), two objective measures (KLD - the Kullback Leibler Divergence and AUC - the Area Under the ROC Curve) and 5 state-of-the-art saliency models (3 acting in spatial domain and 2 acting in compressed domain). The applicative validation is carried out by integrating the MPEG-4 AVC saliency map into a robust watermarking application. As an overall conclusion, the paper demonstrates that although the MPEG-4 AVC standard does not explicitly relies on any visual saliency principle, its stream syntax elements preserve this property. Four main benefits for the MPEG-4 AVC based saliency extraction are thus brought to light: (1) it outperforms (or, at least, is as good as) state-of-the-art uncompressed domain methods, (2) it allows significant gains to be obtained in watermarking transparency (for prescribed data payload and robustness), (3) it is less sensitive to the randomness in the processed visual content, and (4) it has a linear computational complexity. For instance, the ground truth results exhibit absolute relative gains between 60% and 164% in KLD, between 17% and 21% in AUC, and relative gains in KLD sensitivity between 1.18 and 6.12 and in AUC sensitivity between 1.06 and 33.7; the applicative validation brings to light transparency gains up to 10 dB in PSNR
SIMOENS Pieter, JOVESKI Bojan, GARDENGHI Ludovico, MARSHALL Jamie, VANKEIRSBILCK Bert, MITREA Mihai, PRETEUX Francoise, DE TURCK Filip, DHOEDT Bart
Optimized mobile thin clients through an MPEG-4 BiFS semantic remote display framework. Multimedia tools and applications, november 2012, vol. 61, n° 2, pp. 447-470
abstract
According to the thin client computing principle, the user interface is physically separated from the application logic. In practice only a viewer component is executed on the client device, rendering the display updates received from the distant application server and capturing the user interaction. Existing remote display frameworks are not optimized to encode the complex scenes of modern applications, which are composed of objects with very diverse graphical characteristics. In order to tackle this challenge, we propose to transfer to the client, in addition to the binary encoded objects, semantic information about the characteristics of each object. Through this semantic knowledge, the client is enabled to react autonomously on user input and does not have to wait for the display update from the server. Resulting in a reduction of the interaction latency and a mitigation of the bursty remote display traffic pattern, the presented framework is of particular interest in a wireless context, where the bandwidth is limited and expensive. In this paper, we describe a generic architecture of a semantic remote display framework. Furthermore, we have developed a prototype using the MPEG-4 Binary Format for Scenes to convey the semantic information to the client. We experimentally compare the bandwidth consumption of MPEG-4 BiFS with existing, non-semantic, remote display frameworks. In a text editing scenario, we realize an average reduction of 23% of the data peaks that are observed in remote display protocol traffic.
BRILLET Pierre Yves, FETITA Catalin, CAPDEROU André, MITREA Mihai, DREUIL Serge, SIMON Jean-Marc, PRETEUX Francoise, GRENIER Philippe
Variability of bronchial measurements obtained by sequential CT using two computer-based methods. European radiology, may 2009, vol. 19, n° 5, pp. 1139-1147
abstract
This study aimed to evaluate the variability of lumen (LA) and wall area (WA) measurements obtained on two successive MDCT acquisitions using energy-driven contour estimation (EDCE) and full width at half maximum (FWHM) approaches. Both methods were applied to a database of segmental and subsegmental bronchi with LA > 4 mm2 containing 42 bronchial segments of 10 successive slices that best matched on each acquisition. For both methods, the 95% confidence interval between repeated MDCT was between 1.59 and 1.5 mm2 for LA, and 3.31 and 2.96 mm2 for WA. The values of the coefficient of measurement variation (CV10, i.e., percentage ratio of the standard deviation obtained from the 10 successive slices to their mean value) were strongly correlated between repeated MDCT data acquisitions (r > 0.72; p < 0.0001). Compared with FWHM, LA values obtained using EDCE were higher for LA < 15 mm2, whereas WA values were lower for bronchi with WA < 13 mm2; no systematic EDCE underestimation or overestimation was observed for thicker-walled bronchi. In conclusion, variability between CT examinations and assessment techniques may impair measurements. Therefore, new parameters such as CV10 need to be investigated to study bronchial remodeling. Finally, EDCE and FWHM are not interchangeable in longitudinal studies.
Communication dans une conférence à comité de lecture (32)
BEN SAIED Rania, MITREA Mihai
L'impact des étiquettes sémantiques dans l'évaluation subjective de la qualité des séquences vidéo 2D et stéréoscopiques. TAIMA 2018 : Traitement et Analyse de l'Information Méthodes et Applications, Hammamet : 30 avril - 05 mai 2018, Hammamet, Tunisie, 2018
abstract
Cette contribution continue notre étude antérieure vouée à l'évaluation quantitative de l'influence sémantique des étiquettes utilisées dans les tests subjectifs de qualité visuelle. Pour ce faire, nous considérons une approche statistique basée sur : (1) l'estimation de la loi de probabilité des scores accordés par les observateurs humains, (2) des transformations non-linéaires de variable aléatoires reliant les évaluations sur des échelles continues à celles sur des échelles discrètes, et (3) le test binomial. D'un point de vue expérimental, nous considérons les normes ITU-R BT 500-11 et ITU-R BT 1788 pour l'évaluation du contenu vidéo stéréoscopique et respectivement 2D et la méthode d'évaluation continue à un seul stimulus SSCQE (Single Stimulus Continuous Quality Evaluation). Nous traitons 4 corpus vidéo d'une longueur de 20 minutes chacun qui sont évalués par quatre panels de 110 observateurs humains chacun. Les résultats obtenus non seulement permettent de quantifier cette influence sémantique (via la définition et le calcul d'un coefficient d'impact) mais mettent également en évidence d'autres facteurs subjectifs différentiant la perception de qualité du contenu stéréoscopique par rapport au contenu 2D
BEN SAIED Rania, MITREA Mihai
The Gaussian hypothisis in subjective quality evaluation for stereoscopic and 2D video content. EUVIP 2018 : 7th European Workshop on Visual Information Processing, Los Alamitos : IEEE Computer Society, 26-28 november 2018, Tampere, Finland, 2018, pp. 1-6, ISBN 978-1-5386-6897-9
abstract
The present paper investigates two controversial issues related to subjective video quality assessment, namely the relationship between continuous and discrete scale evaluations and the impact of the Gaussian hypothesis in modeling the scores assigned by the human observers. The theoretical background is obtained by reconsidering a non-linear random variable transformation formula established in our previous work and by instantiating it for both Gaussian and non-Gaussian cases. The experimental results are obtained by considering a total of 240 human observers, evaluating both stereoscopic and 2D video quality content, each of which considered at both high quality and low quality (as a priori expressed by objective measures). The results demonstrate that the Gaussian hypothesis cannot be accepted and that the relative error in the mean opinion score estimation can be reduced by 2-4% when considering a non- Gaussian model
BOUJELBANE Ismail, MITREA Mihai
Vers la virtualisation des objets connectés personnels. TAIMA 2018 : Traitement et Analyse de l'Information Méthodes et Applications, Hammamet : 30 avril - 05 mai 2018, Hammamet, Tunisie, 2018
abstract
Notre étude porte sur l'une des limitations aujourd'hui courantes dans le développement des solutions logicielles pour les objets connectés personnels (wearables) : la coexistence d'une multitude de formats d'échange de données entre l'objet et l'application qui l'utilise. En outre, malgré le foisonnement des normes de jure sous-jacents, ces formats restent assez souvent propriétaires et tenus confidentiels par leur fournisseur : les choix du développeur sont ainsi contraints à l'utilisation de certaines APIs propriétaires et, implicitement, à certains logiciels adossés à celles-ci. Pour s'affranchir de cette limitation, nous spécifions un nouveau type de composant, nommé " connecteur virtuel " permettant de convertir en temps réel des données propriétaires générées par l'objet dans un format ouvert. Constitué par trois sous-modules (le " connecteur ", l'" extracteur de données " et le " gestionnaire de format "), ce module est évalué dans un cas d'usage lié à la virtualisation des bracelets connectés
The present paper goes one step further in our study devoted to the assessment of the semantic impact of the label considered in the subjective evaluation tests for video content. In this respect, we consider a statistical approach based on: (1) probability law estimation for scores assigned by the human observers, (2) non-linear random variables transforms connecting continuous and discrete-scale evaluations, and (3) the binomial test on probability. The experiments consider the ITU-R BT 500-11 and ITU-R BT 1788 recommendations and the SSCQE (Single Stimulus Continuous Quality Evaluation), for assessing 2D and stereoscopic video, respectively. Four video corpora of 20 minutes each have been scored by four panels of 110 observers each. The experimental results brought to light the quantitative impact of the labels and pointed to some other subjective discriminative factors between 2D and stereoscopic video quality evaluation.
AMMAR Marwa, MITREA Mihai, BOUJELBANE Ismail
HEVC stream saliency extraction: synergies between FIT and information theory principles. IPTA 2017 : 7th International Conference on Image Processing Theory, Tools and Applications, Los Alamitos : IEEE Computer Society, 28 november - 01 december 2017, Montreal, Canada, 2017, pp. 1-5, ISBN 978-1-5386-1842-4
abstract
The present paper studies the potential synergies between a popular approach in saliency extraction - FIT (feature integration theory) and source coding principles. By combining these two approaches, a new saliency model, extracted directly at the HEVC steam syntax elements level is defined. The experiments confront the new model to the human saliency, captured by eye-tracking devices. They consider a reference corpus representing density fixation maps, two objective criteria, two objective measures and 7 state-of-the-art saliency models (3 acting in pixel domain and 4 acting in compressed domain)
BENOIS-PINEAU Jenny, MITREA Mihai
Extraction of saliency in images and video: problems, methods and applications: a survey. IPTA 2017 : 7th International Conference on Image Processing Theory, Tools and Applications, Los Alamitos : IEEE Computer Society, 28 november - 01 december 2017, Montreal, Canada, 2017, pp. 1-6, ISBN 978-1-5386-1842-4
abstract
Rather than meeting the theoretical, methodological and applicative expectancies, the impressing number of state-of-the-art saliency oriented studies raises new fundamental questions about the very nature of this psycho-cognitive process. Such questions encompass fundamental modeling aspects from the very saliency dependency on the representation format to its potential relationship to other fundamental research areas, like information theory, for instance. The present survey, structured according to three main saliency applicative fields (visual quality evaluation, watermarking and task-oriented computer vision) is meant to identify the latest trends of research
MAHÉ Gael, MOISAN Lionel, MITREA Mihai
An image-inspired audio sharpness index. EUSIPCO 2017 : 25th European Signal Processing Conference, Los Alamitos : IEEE Computer Society, 28 august - 02 september 2017, Kos Island, Greece, 2017, pp. 683-687, ISBN 978-0-9928626-7-1
abstract
We propose a new non-intrusive (reference-free) objective measure of speech intelligibility that is inspired from previous works on image sharpness. We define the audio Sharp- ness Index (aSI) as the sensitivity of the spectrogram sparsity to the convolution of the signal with a white noise, and we calculate a closed-form formula of the aSI. Experiments with various speakers, noise and reverberation conditions show a high correlation between the aSI and the well-established Speech Transmission Index (STI), which is intrusive (full-reference). Additionally, the aSI can be used as an intelligibility or clarity criterion to drive sound enhancement algorithms. Experimental results on stereo mixtures of two sounds show that blind source separation based on aSI maximization performs well for speech and for music
AMMAR Marwa, MITREA Mihai, BOUJELBANE Ismail, LE CALLET Patrick
HEVC saliency map computation. ELECTRONIC IMAGING 2016 : Human Vision and Electronic Imaging XX, Springfield : Society for Imaging Science and Technology, 15-18 february 2016, San Francisco, United States, 2016, pp. HVEI-107-1-8, ISBN 2470-1173
URL: http://ist.publisher.ingentaconnect.com/contentone/ist/ei/2016/00002016/00000016/art00031
abstract
This paper investigates whether the information related to the human visual saliency is still preserved at the level of the HEVC compressed stream syntax elements. In this respect, a new saliency model, matched to the peculiarities of this emerging standard is defined. It consists of four elementary maps, describing the four main saliency features: intensity, color, orientation and motion. These maps are defined based on the energies of the luma and chroma coefficients, on the variations of the intra prediction modes and on the energy of motion vectors, respectively. They are fusioned according to 48 static and static-dynamic pooling formulas. The results are compared to three state-of-the-art uncompressed (pixel) domain as well as to the MPEG-4 AVC compressed domain saliency maps. It is brought to light that the HEVC saliency model outperforms (with singular exceptions) the state-of-the-art uncompressed domain and is as good as MPEG-4 AVC saliency model. We can thus state that, as its MPEG-4 AVC ancestor, although not designed based upon visual saliency principles, the HEVC compression standard preserves this human visual property at the level of its syntax elements
BEN SAIED Rania, MITREA Mihai
Assessing the impact of the semantic labels in subjective video quality evaluation. 11th IMA International Conference on Mathematics in Signal Processing, 12-14 december 2016, Birmingham, United Kingdom, 2016, pp. 16-20
abstract
The present study quantifies the bias induced in subjective video quality experiments by the semantic labels (e.g. "Bad", "Poor", "Fair", "Good", and "Excellent") generally associated to the grading scale. In this respect, a common theoretical ground for subjective quality evaluation, encompassing both continuous and discrete scale evaluations is deployed. The experimental study is performed according to the ITU-R BT 500-11 and to the SSCQE (Single Stimulus Continuous Quality Evaluation) method. A total of 330 human observers are inquired. They are grouped in three main panels of 110 observers, on panel for each type of content under investigation (HD 3D TV, HD TV, low quality video). In order to grant statistical relevance, each main panel was split into three sub-panels, referred to as the reference (60 observers), validation (25 observers) and cross-checking (25 observers) panels. The experimental results demonstrate that the "Excellent" label has an important impact (biases up to 65% with respect to the a priori expected situation), independent with respect to the content under investigation. Large biases are also brought to light for other semantic labels; however, these seem not to be solely induced by the label semantic
GANJI Rama Rao, MITREA Mihai, PANOVSKI Dancho, JOVESKI Bojan
Improving the RDP based applications by using HTML5 content representation. ELECTRONIC IMAGING 2016 : Mobile Devices and Multimedia : Enabling Technologies, Algorithms, and Applications, Springfield : Society for Imaging Science and Technology, 14-18 february 2016, San Francisco, United States, 2016, pp. MOBMU-293-1-7, ISBN 2470-1173
URL: http://www.ingentaconnect.com/contentone/ist/ei/2016/00002016/00000007/art00011
abstract
Despite the large variety of "off the shelf" solutions and academic research studies, application virtualization for cloud distribution is still an open research topic, with significant issues to be solved, ranging from bridging the gap between users expectation of simple, intuitive and user-friendly access from any type of terminal and the fragmented landscape of SaaS offer, with peculiarities related to the hardware/software configurations and the optimization of the technical resources consumption. Our study investigates the possibility of using HTML5 a virtualization tool for RDP-based applications. Architectural modules related to the RDP content interception, conversion, adaptation, remote rendering and interaction are specified, designed and implemented. This architecture is validated under the framework of the MEDUSA European project, in partnership with medical institutions. The testbed considers a server and 5 mobile users, with heterogeneous devices (tablets, smartphones, laptops) running under iOS, Android and Windows operating systems. The objective/subjective evaluations demonstrated that: (1) the user experience is not reduced by the virtualization, (2) the network consumption is reduced by a factor of 1.8 with respect to state-of-the-art solutions
AMMAR Marwa, MITREA Mihai, HASNAOUI Marwen, LE CALLET Patrick
Visual saliency in MPEG-4 AVC video stream. IST & SPIE Electronic Imaging 2015 : Human Vision and Electronic Imaging XX, Bellingham; Springfield : SPIE-IS&T, 09-12 february 2015, San Francisco, United States, 2015, pp. 93940X, ISBN 978-1-6284-1484-4
abstract
Visual saliency maps already proved their efficiency in a large variety of image/video communication application fields, covering from selective compression and channel coding to watermarking. Such saliency maps are generally based on different visual characteristics (like color, intensity, orientation, motion,…) computed from the pixel representation of the visual content. This paper resumes and extends our previous work devoted to the definition of a saliency map solely extracted from the MPEG-4 AVC stream syntax elements. The MPEG-4 AVC saliency map thus defined is a fusion of static and dynamic map. The static saliency map is in its turn a combination of intensity, color and orientation features maps. Despite the particular way in which all these elementary maps are computed, the fusion techniques allowing their combination plays a critical role in the final result and makes the object of the proposed study. A total of 48 fusion formulas (6 for combining static features and, for each of them, 8 to combine static to dynamic features) are investigated. The performances of the obtained maps are evaluated on a public database organized at IRCCyN, by computing two objective metrics: the Kullback-Leibler divergence and the area under curve
GANJI Rama Rao, MITREA Mihai, JOVESKI Bojan, PANOVSKI Dancho
Cross-standard user description in mobile, medical oriented virtual collaborative environments. IST & SPIE Electronic Imaging 2015 : Mobile Devices and Multimedia : Enabling Technologies, Algorithms, and Applications, Bellingham; Springfield : SPIE-IS&T, 10-11 february 2015, San Francisco, United States, 2015, pp. 1-13, ISBN 978-1-6284-1501-8
abstract
By combining four different open standards belonging to the ISO/IEC JTC1/SC29 WG11 (a.k.a. MPEG) and W3C, this paper advances an architecture for mobile, medical oriented virtual collaborative environments. The various users are represented according to MPEG-UD (MPEG User Description) while the security issues are dealt with by deploying the WebID principles. On the server side, irrespective of their elementary types (text, image, video, 3D, …), the medical data are aggregated into hierarchical, interactive multimedia scenes which are alternatively represented into MPEG-4 BiFS or HTML5 standards. This way, each type of content can be optimally encoded according to its particular constraints (semantic, medical practice, network conditions, etc.). The mobile device should ensure only the displaying of the content (inside an MPEG player or an HTML5 browser) and the capturing of the user interaction. The overall architecture is implemented and tested under the framework of the MEDUSA European project, in partnership with medical institutions. The testbed considers a server emulated by a PC and heterogeneous user devices (tablets, smartphones, laptops) running under iOS, Android and Windows operating systems. The connection between the users and the server is alternatively ensured by WiFi and 3G/4G networks
GARBOAN Adriana, MITREA Mihai, PRETEUX Francoise
Video fingerprinting for second screen synchronization. EuroITV 2013 : 11th European Interactive TV Conference, Multiscreen 2013 workshop, 24-24 june 2013, Como, Italy, 2013
abstract
The present paper advances a robust video fingerprinting method to be deployed within second screen synchronization applications. The novelty consists in tracking camera recorded video sequences based on their visual content. In this respect, the advanced system creates synergies between two architectural modules, a bag of visual words framework ensuring robust and scalable query localization in the reference database and an aggregation block providing an accurate synchronization of the video sequences as well as the system's decision. The system was tested on a reference database of 14 hours of video content and on a query dataset of 5 hours obtained by the StirMark geometric random bending and by live camera recording. The evaluation resulted into average false alarm rate of 1.3×10 5, probability of missed detection of 0.05 and F1 score equal to 0.96
BELHAJ ABDALLAH Maher, MITREA Mihai, PRETEUX Francoise
MPEG-4 AVC perceptual masking. Case study: Robust video watermarking . ISCE 2011 : 15th IEEE International Symposium on Consumer Electronics, IEEE, 14-17 june 2011, Singapore, Singapore, 2011, pp. 598-601, ISBN 978-1-61284-843-3
abstract
Since the pioneering studies of Peterson and Watson, the perceptual masking for raw image has been playing an always increasing role in image processing: compression, filtering, indexing, watermarking, fingerprinting are just examples of application fields exploiting the human visual system peculiarities in order to increase their practical impact. The present study establishes the first accurate perceptual masking model addressing the MPEG-4 AVC domain peculiarities. In other words, we computed the MPEG-4 AVC visibility thresholds, i.e. the maximum amount of additive noise which can imperceptibly affect the quantized values of the intra prediction errors. This new masking model has been tested against other masking models, under the watermarking framework. On the one hand, when imposing a prescribed robustness and a fixed data payload, significant gains in transparency were obtained: the PSNR is increased by 2dB while the Watson's DVQ (Digital Video Quality) is decreased by a half. On the other hand, when imposing a prescribed robustness and transparency, gains in data payload by about 5% are reached
CHAMMEM Afef, MITREA Mihai, PRETEUX Francoise
DWT-based stereoscopic image watermarking. SPIE Photonics West 2011 : Stereoscopic Displays and Applications XXII, SPIE, 24-27 january 2011, San Francisco, United States, 2011, vol. 7863, ISBN 978-0-81948-400-0
abstract
Watermarking already imposed itself as an effective and reliable solution for conventional multimedia content protection (image/video/audio/3D). By persistently (robustly) and imperceptibly (transparently) inserting some extra data into the original content, the illegitimate use of data can be detected without imposing any annoying constraint to a legal user. The present paper deals with stereoscopic image protection by means of watermarking techniques. That is, we first investigate the peculiarities of the visual stereoscopic content from the transparency and robustness point of view. Then, we advance a new watermarking scheme designed so as to reach the trade-off between transparency and robustness while ensuring a prescribed quantity of inserted information. Finally, this method is evaluated on two stereoscopic image corpora (natural image and medical data).
GANJI Rama Rao, BELHAJ ABDALLAH Maher, MITREA Mihai, PRETEUX Francoise
MPEG-4 AVC re-encoding for watermarking purposes. WoSSPA '11 : The 7th International Workshop on Systems, Signal Processing and their Applications, IEEE, 09-11 may 2011, Tipaza, Algeria, 2011, pp. 319-322, ISBN 978-1-4577-0689-9
abstract
The present paper deals with robust and transparent video watermarking in the MPEG-4 AVC domain. It reconsiders and extends an insertion technique previously presented by the authors in order to increase its overall performances (transparency and robustness against re-encoding attacks). In this respect, an objective study on the practical impact of the errors induced by the MPEG-4 AVC (iterative) decoding/re-encoding is carried out. The experiments are performed on the MEDIEVALS corpus (more than 1h of video). The results show an increase of the PSNR of about 1.3dB and a decrease of the BER by about 10%.
GARBOAN Adriana, MITREA Mihai, PRETEUX Francoise
DWT-based robust video fingerprinting. EUVIP '11 : 3rd European Workshop on Visual Information Processing, IEEE, 04-06 july 2011, Paris, France, 2011, pp. 216-221, ISBN 978-1-4577-0072-9
abstract
Video fingerprints are short features extracted from a video sequence in order to uniquely identify that visual content and its replicas. The present paper introduces a new DWT-based robust video fingerprinting method. In this respect, a characteristic set of wavelet coefficients is used as fingerprint. Due to the inner DWT properties with respect to content preserving attacks (such as linear filtering, sharpening, geometric, conversion to grayscale, small rotations, contrast changes), good results in terms of probability of missed detection and probability of false alarm are obtained. The video corpus consists of 3 hours of heterogeneous original content and of its attacked versions (a total of 21 hours of video content).
GARBOAN Adriana, MITREA Mihai, PRETEUX Francoise
Video retrieval by means of robust fingerprinting. ISCE '11 : The 15th IEEE International Symposium on Consumer Electronics, IEEE, 14-17 june 2011, Singapore, Singapore, 2011, pp. 299-303, ISBN 978-1-61284-843-3
abstract
Uniquely identifying visual content remains a challenging issue for a large variety of nowadays applications, as video browsing, database search and multimedia security, for instance. In this respect, our study brought to light a simple yet efficient fingerprinting technique allowing short video sequences to be tracked. Three corpora, all of them containing 3780 video excerpts, with different excerpts lengths (20 seconds, 40 seconds and 60 seconds) were considered in the experiments. The quantitative results established that the average probability errors for both missed detection and false alarm are lower than 0.0007. These good practical results derive from the very fine mathematical properties of stationarity governing the DWT coefficients representing the fingerprint.
HASNAOUI Marwen, BELHAJ ABDALLAH Maher, MITREA Mihai, PRETEUX Francoise
MPEG-4 AVC stream watermarking by m-QIM techniques. SPIE Photonics West 2011 : Multimedia on Mobile Devices and Multimedia Content Access : Algorithms and Systems V, Bellingham;Springfield : SPIE;IS&T Electronic Imaging, 22-27 january 2011, San Francisco, United States, 2011, vol. 7881, pp. 78810L:1-78810L:11, ISBN 978-0-81948-418-5
abstract
The present paper is devoted to the MPEG-4 AVC/H.264 video stream protection by means of watermarking techniques. The embedding process is carried out in quantized index domain and relies on the m-QIM (m-arry Quantisation Index Modulation) principles. In order to cope with the MPEG-4 AVC peculiarities, the Watson's perceptual model is reconsidered and discussed. The experimental results correspond to the MEDIEVALS (a French National Project) corpus of 4 video sequences of about 15 minutes each, encoded at 512kbps. The transparency is assessed by both subjective and objective measures. The transcoding (down to 64kbps) and geometric (StirMark) attacks result in BER of 6.75% and 11.25%, respectively. In order to improve robustness, an MPEG-4 AVC syntax-driven counterattack is considered: this way, the two above mentioned attacks lead to BER of 2% and 10%, respectively. Finally, the overall theoretical relevance of these results is discussed by estimating the related channel capacities.
JOVESKI Bojan, GARDENGHI Ludovico, MITREA Mihai, PRETEUX Francoise
Towards collaborative MPEG-4 BiFS mobile thin remote viewer. ISCE '11 : The 15th IEEE International Symposium on Consumer Electronics, IEEE, 14-17 june 2011, Singapore, Singapore, 2011, pp. 107-111, ISBN 978-1-61284-843-3
abstract
The proof of concepts (PoC) for MPEG-4 BiFS & LASeR mobile thin remote viewer has been already obtained. The present paper goes one step further, from PoC towards applications supporting multiuser collaboration in virtual environments. In this respect, the previous architecture is reconsidered and enriched with new components. This way, significant gains with respect to competitor remote viewer solutions (HEXTILE optimised VNC) are obtained. First, by empowering the basic BiFS mechanisms with lossy and lossless compression, the downlink bandwidth consumption is reduced by an average factor of 2. Secondly, by introducing a new architectural component (Interaction Manager), the uplink consumption corresponding to the user interaction is reduced by an average factor of 10, while the roundtrip network time copes with the real time requirements (lower than 30 ms in Wi-Fi setups). Finally, the CPU consumption is reduced by an average factor of 6. The above-mentioned quantitative results correspond to two use cases, namely text editing (gEdit) and www browsing (Epiphany).
JOVESKI Bojan, SIMOENS Pieter, GARDENGHI Ludovico, MARSHALL Jamie, MITREA Mihai, VANKEIRSBILCK Bert, PRETEUX Francoise, DHOEDT Bart
Towards a multimedia remote viewer for mobile thin clients. SPIE Photonics West 2011 : Multimedia on Mobile Devices and Multimedia Content Access : Algorithms and Systems V, SPIE;IS&T Electronic Imaging : Bellingham;Springfield, 22-27 june 2011, San Francisco, United States, 2011, vol. 7881, pp. 788102:1-788102:9, ISBN 978-0-81948-418-5
[PDF]
abstract
Be there a traditional mobile user wanting to connect to a remote multimedia server. In order to allow them to enjoy the same user experience remotely (play, interact, edit, store and share capabilities) as in a traditional fixed LAN environment, several dead-locks are to be dealt with: (1) a heavy and heterogeneous content should be sent through a bandwidth constrained network; (2) the displayed content should be of good quality; (3) user interaction should be processed in real-time and (4) the complexity of the practical solution should not exceed the features of the mobile client in terms of CPU, memory and battery. The present paper takes this challenge and presents a fully operational MPEG-4 BiFS solution
BELHAJ ABDALLAH Maher, MITREA Mihai, PRETEUX Francoise, DUTA Sorin
MPEG-4 AVC robust video watermarking based on QIM and perceptual masking . COMM '10 : The 8th International Conference on Communications, IEEE, 10-12 june 2010, Bucharest, Romania, 2010, pp. 477-480, ISBN 978-1-4244-6360-2
abstract
The present paper advances a new watermarking method featuring robustness against noise addition, transcoding and the StirMark (geometric) attacks. The mark is inserted into the MPEG-4 AVC quantisation indexes selected according to an energy-based selection criterion validated by information theory basic concepts. The insertion procedure combines the QIM principles and a perceptual mask obtained by adapting the popular Watson results. The video corpus consists of 7 video sequences (2 hours). The transparency is exhibited by both subjective and objective evaluations.
HASNAOUI Marwen, BELHAJ ABDALLAH Maher, MITREA Mihai, PRETEUX Francoise
MPEG-4 AVC stream watermarking by ST-mDM techniques electronics, circuits, and systems. ICECS '10 : 17th IEEE International Conference on Electronics, Circuits, and Systems, IEEE, 12-15 december 2010, Athens, Greece, 2010, pp. 487-490, ISBN 978-1-4244-8155-2
abstract
The present paper introduces a new insertion/detection technique for MPEG-4 AVC stream protection by means of watermarking techniques. The embedding process relies on the ST-DM (Spread Transform - Dither Modulation) principles and adapts/extends the Watson's perceptual model so as to cope with the compressed domain peculiarities. In contrast to the state-of-the art techniques considering only binary insertion techniques, our paper demonstrates the first m-ary DM technique; in such a way, for a prescribed transparency and robustness, the data payload is increased by a factor of log2m. The experimental results correspond to the MEDIEVALS (French National Project) corpus of 4 video sequences of about 15 minutes each. The robustness against noise addition, transcoding and geometric (StirMark) attacks is proved. The transparency is assessed by both subjective and objective measures.
JOVESKI Bojan, MITREA Mihai, PRETEUX Francoise
MPEG-4 LASeR - based thin client remote viewer. EUVIP '10 : The 2nd European Workshop on Visual Information Processing, IEEE, 05-06 july 2010, Paris, France, 2010, pp. 125-128, ISBN 978-1-4244-7288-8
abstract
Nowadays, to implement a remote display for mobile thin clients is a hot research topic: specufying a high-performing compression algorithm for heterogeneous content (text, graphics, image, video, 3D, ...) and ensuring versatile and user-friendly interaction are two constraints that should be jointly fulfilled. This paper establishes the appropriateness of the LASeR (MPEG-4 Part 20) technologies for such an application. In this respect, a software architecture based on LASeR is first presented. Secondly, a study on the corresponding user visual experience is reported. Finally, the bandwidth consumption of this solution is assed and compared to its competitors, be they from the wired (VNC) or wireless (BiFS) frameworks.
DUMITRU Corneliu Octavian, MITREA Mihai, PRETEUX Francoise
Theoretical limits in DWT video watermarking. Mathematics of data/image pattern recognition, compression, and encryption with applications XI 2008, Bellingham, Wash. : SPIE, 12-13 august 2008, San Francisco, United States, 2008, vol. 7075, pp. 70750C-1-70750C-9, ISBN 978-0-8194-7295-3
abstract
Nowadays, nobody doubts about the huge economical benefits the watermarking solutions will one day bring. The paper is devoted to the theoretical evaluation of the watermarking capacity, i.e. devoted to find out with mathematical rigour the maximum amount of information which can be inserted into the DWT of natural video, for prescribed constraints of transparency and robustness. The starting point is the accurate statistical model for the watermarking attacks the authors already reported. In this paper, in addition to the classical Shannon solutions, the capacity is evaluated by two approaches: (1) a method developed in order to increase speed and precision for watermarking evaluations and (2) the general Blahut Arimoto algorithm, adapted by Justin Dauwels for the discrete case. The experiments are run on a video corpus of 10 video sequences of about 25 minutes each
DUMITRU Corneliu Octavian, MITREA Mihai, PRETEUX Francoise
DCT domain video watermarking : attack estimation and capacity evaluation. ICINCO 2008 : 5th International Conference on Informatics in Control, Automation and Robotics, Scitepress, 11-15 may 2008, Funchal, Portugal, 2008, pp. 239-244, ISBN 978-989-8111-31-9
abstract
The first difficulty when trying to evaluate with accuracy the video watermarking capacity is the lack of a reliable statistical model for malicious attacks. The present paper brings into evidence that the attack effects in the DCT domain are stationary and computes the corresponding pdfs. In this respect, an in-depth statistical approach is deployed by combining Gaussian mixture estimation with the probability confidence limits. Further on, these pdfs are involved in capacity computation. The experimental results are obtained on a corpus of 10 video sequences (avout 25 minutes each), with heterogeneous content
DUMITRU Corneliu Octavian, MITREA Mihai, PRETEUX Francoise
Wavelet-based video modelling. ELMAR 2008 : 50th International Symposium, IEEE, 10-12 september 2008, Zadar, Croatia, 2008, pp. 105-108, ISBN 978-1-4244-3364-3
abstract
Regardless the targeted application (compression, watermaking, texture analysis, indexation, …), image/video modelling in the DWT domain is generally approached by tests of concordance with some well known pdfs (like Gaussian, generalised Gaussian or Laplace, for instance). Instead of forcing the images/videos to stick to such theoretical models, our study aims at estimating their inner statistical behaviour. In this respect, we first prove that a law modelling the video DWT coefficients decreasingly ordered does exist. Then, we estimate this law by Gaussian mixtures and finally we identify the generality of such model with respect to the data on which it was computed and to the estimation method it relies on. The usefulness of such model was checked-up under the framework of watermarking and indexation applications developed in industrial partnership at the ARTEMIS Department
DUMITRU Corneliu Octavian, MITREA Mihai, PRETEUX Francoise, PATHAK A.
Probability density function estimation for video in the DCT domain. Image processing : algorithms and systems VI 2008, Bellingham : SPIE, 28-29 january 2008, San Jose, United States, 2008, vol. 6812, pp. 68120L , ISBN 978-0-8194-6984-7
abstract
Regardless the final targeted application (compression, watermarking, texture analysis, indexation...), image/video modelling in the DCT domain is generally approached by tests of concordance with some well known pdfs (like Gaussian, generalised Gaussian, Laplace, Rayleigh...). Instead of forcing the images/videos to stick to such theoretical models, our study aims at estimating the true pdf characterising their behaviour. In this respect, we considered three intensively used ways of applying DCT, namely on whole frames, on 4x4 blocks, and on 8x8 blocks. In each case, we first prove that a law modelling the corresponding coefficients exists. Then, we estimate this law by Gaussian mixtures and finally we identify the generality of such model with respect to the data on which it was computed and to the estimation method it relies on
DUTA Sorin, MITREA Mihai, BELHAJ ABDALLAH Maher, PRETEUX Francoise
A comparative study on insertion strategies in MPEG-4 AVC watermarking. Mathematics of data/image pattern recognition, compression, and encryption with applications XI 2008, Bellingham, Wash. : SPIE, 12-13 august 2008, San Diego, United States, 2008, vol. 7075, pp. O9:1-O9:9, ISBN 978-0-8194-7295-3
abstract
High speed, low complexity, and interoperability are just three of the main advantages turning the MPEG stream watermarking into a hot research topic. Unfortunately, viable solutions (in terms of robustness, data payload and transparency) are yet to be found. In their previous work, the authors computed general models for the watermarking attack effects (StirMark, linear & nonlinear filtering, rotations) in the MPEG 4 AVC stream. These models (expressed as noise matrices) are now the starting point for evaluating three classes of watermarking insertion techniques (substitutive, additive, and multiplicative). For each class, a specific set of noise matrices is first computed by particularising the general model. Secondly, the corresponding capacity values (i.e. the largest data payload which can be inserted for prescribed transparency and robustness) are computed. The paper is concluded with a comparison among these method performances. The experiments are run on a video corpus of 10 video sequences of about 25 minutes each
DUTA Sorin, MITREA Mihai, PRETEUX Francoise, BELHAJ ABDALLAH Maher
MPEG-4 AVC domain watermarking transparency. Mobile Multimedia/Image Processing, Security, and Applications 2008, Bellingham, Wash. : SPIE, 18-20 march 2008, Orlando, United States, 2008, vol. 6982, pp. 69820F, ISBN 978-0-8194-7173-4
abstract
The ever-increasing Internet distribution of video content is echoed in ever-increasing efforts to devise systems balancing copyright protection and user rights. Watermarking is such an example: by persistently and imperceptibly associating some data with the host video, it offers at the same time a reliable and user-friendly solution for copyright infringement tracking. This paper takes a closer look at the apparent contradiction between watermarking (using the visual redundancy of the video to embed the data) and compression (eliminating the visual redundancy in order to speed up distribution and to alleviate storage requirements). In this respect, the viability of compressed domain watermarking is evaluated by analysing the visual effects of the MPEG-4 AVC stream alteration. The corpus consists of 10 video sequences of about 25 minutes each, coded at 256kbps and 64 kbps
DUTA Sorin, MITREA Mihai, PRETEUX Francoise
Capacity evaluation for MPEG-4 AVC watermarking. SPIE Optical and digital image processing 2008 , Bellingham : SPIE, 07-09 april 2008, Strasbourg, France, 2008, vol. 7000, pp. 1-10, ISBN 978-0-8194-7198-7
abstract
Nowadays, robust watermarking clearly identified its functionality within the multimedia production chain, from the content creation to the end-user consumption: property right identification and copy-maker tracking. In the quest for the speed required by today's real-time applications, compressed-domain watermarking becomes a hot research topic. This study evaluates the watermarking capacity in the MPEG-4 AVC domain in order to establish whether and to what extent compressed domain watermarking is viable. In this respect, the additive watermarking techniques are modelled by discrete noisy channels with non-causal side information at the transmitter. The study considers several attacks (linear and non-linear filtering, geometric) and computes the capacity of the corresponding channels. The experimental results are obtained out of processing a natural video corpus of 10 video sequences belonging to different movies, each of them about 25 minutes long (35000 frames in each video sequence)
DUTA Sorin, MITREA Mihai, PRETEUX Francoise, RIFFAUD L.A
The watermarking attacks in the MPEG-4 AVC domain. Image processing : algorithms and systems VI 2008, Bellingham, Wash. : SPIE, 28-29 january 2008, San Jose, United States, 2008, vol. 6812, pp. 68120P, ISBN 978-0-8194-6984-7
abstract
The explosion of VoD and HDTV services opened a new direction in watermarking applications: compressed domain watermarking, promising at least tenfold speed increase. While sound technical approaches to this emerging field are already available in the literature, at our best knowledge the present paper is the first related theoretical study. It considers the ISO/IEC 14496-10:2005 standard (also known as MPEG-4 AVC) and objectively describes with information theory concepts (noisy channel, noise matrices) the effects of the real-life watermarking attacks (like rotations, linear and non-linear filtering, StirMark). All the results are obtained on a heterogeneous corpus of 7 video sequences summing up to about 3 hours
MITREA Mihai, DUMITRU Corneliu Octavian, DUTA Sorin, PRETEUX Francoise, VLAD Adriana
A comparative study on video watermarking capacity. Communications 2008 : 7th International Conference, Bucharest, Romania : Printech, 05-07 june 2008, Bucharest, Romania, 2008, pp. 335-338, ISBN 978-606-521-008-0
abstract
From both theoretical and applicative points of view, estimating with accuracy the video watermarking capacity is an ever challenging research topic. This paper completes the author's previous study and provides reference capacity values for a large variety of real life scenarios: the uncompressed and compressed (MPEG 4 AVC) video, the two most intensively used transforms in watermarking (DCT and DWT), different levels of robustness (from linear and non-linear filtering to geometrics attacks), and transparency (expressed by signal to noise ratios from 25 to 35 dB). The experiments are performed in industrial partnership, out of processing a heterogeneous video corpus of about 4 hours
Autres rapports (8)
JOVESKI Bojan, MITREA Mihai, BOUJELBANE Ismail
MPEG-21-UD for personalisation of real live application . january 2018
abstract
FitBit devices are very popular with more than 5M devices sold by 2016 year. These devices as well as the applications FitBit provides (or that can be developed using FitBit APIs) are proprietary, based on undisclosed data formats. Hence, developing an MPEG-21 UD application for such proprietary data can be considered as a successful validation of the reference software viability
MITREA Mihai
IoMT use case: mThing content authentication with blockchain technologies. april 2018
abstract
Assume the case of a face recognition application: multiple cameras that exchange information with a remote processing unit are meant to track a suspicious person. However, before making a decision for a physical intervention, the content exchanged among these mThings should be registered into a trusted database
MITREA Mihai, JOVESKI Bojan
MPEG-UD vs. GDPR. april 2018
abstract
As MPEG-UD describes users data, with this input we want to analyze the ability of the MPEG-UD standard to respond to the new regulation of using the data imposed by the European union. EU General Data Protection Regulation (GDPR) [1] defines new principles that organizations must follow when collecting, processing and storing individuals' personal data
MITREA Mihai, JOVESKI Bojan
The usage of block-chain solutions for MPEG-UD data: monetizing, tracking, authentication. april 2018
abstract
Block-chain becomes the world's leading solution for digital assets. It's based on large production block chains platform for building better systems especially for monetizing, tracking and authentication purposes. In this respect, using the block-chain becomes essential part of any new application so we are submitting a new use case to describe its usage
MITREA Mihai, PREDA Marius
Updated requirements for Internet of Media Things. april 2018
abstract
The Internet of Media Things and Wearables (IoMT & W) is the collection of interfaces, protocols and associated media-related information representations that enable advanced services and applications based on human to device and device to device interaction, in physical and virtual environments. Information refers to data sensed and processed by a device, and/or communicated to a human or another device
MITREA Mihai, PREDA Marius
On the usage of blockchain solutions for IOMT. january 2018
abstract
During the 120th meeting held in Macao, MPEG issued an output document w17250 - Ideas on using digital coins for IoMT. The basic idea is that IoMT eco-system can only grow if it conveys economic value. While IoTs in general has value, media processing performed by IoMTs has an even greater value. Hence, the MPEG interest in using digital coins for facilitating the exchanges between media related devices and for supporting media related processing was thus confirmed
MITREA Mihai, PREDA Marius
On the usage of IOMT in real life applications. october 2017, ISBN m41828
abstract
MPEG 120 th Meeting, 23 October 2017 to Friday, 27 October 2017 , Macau, China. Report ISO/IEC JTC 1/SC 29/WG 11 N 17203
KIM Sang Kyun, MITREA Mihai, PREDA Marius, CHIARIGLIONE Leonardo
AHG report on Media-centric Internet of Things (MIoT) and MPEG wearable. june 2015, ISBN m36419
abstract
112th MPEG meeting, Warsaw, PL - June 2015 - ISO/IEC JTC 1/SC 29/WG 11 N15327