Multimedia Security 1: Authentication and Data Hiding

ISBN-13: 9781789450262 / Angielski / Twarda / 2022 / 320 str.

William Puech

Spis treści

Foreword by Gildas Avoine xiForeword by Cédric Richard xiiiPreface xvilliam PUECHChapter 1 How to Reconstruct the History of a Digital Image, and of Its Alterations 1Quentin BAMMEY, Miguel COLOM, Thibaud EHRET, Marina GARDELLA, Rafael GROMPONE, Jean-Michel MOREL, Tina NIKOUKHAH and Denis PERRAUD1.1 Introduction 21.1.1 General context 21.1.2 Criminal background 31.1.3 Issues for law enforcement 41.1.4 Current methods and tools of law enforcement 51.1.5 Outline of this chapter 51.2 Describing the image processing chain 81.2.1 Raw image acquisition 81.2.2 Demosaicing 81.2.3 Color correction 101.2.4 JPEG compression 111.3 Traces left on noise by image manipulation 111.3.1 Non-parametric estimation of noise in images 111.3.2 Transformation of noise in the processing chain 131.3.3 Forgery detection through noise analysis 151.4 Demosaicing and its traces 181.4.1 Forgery detection through demosaicing analysis 191.4.2 Detecting the position of the Bayer matrix 201.4.3 Limits of detection demosaicing 231.5 JPEG compression, its traces and the detection of its alterations 231.5.1 The JPEG compression algorithm 231.5.2 Grid detection 251.5.3 Detecting the quantization matrix 271.5.4 Beyond indicators, making decisions with a statistical model 281.6 Internal similarities and manipulations 311.7 Direct detection of image manipulation 331.8 Conclusion 341.9 References 35Chapter 2 Deep Neural Network Attacks and Defense: The Case of Image Classification 41Hanwei ZHANG, Teddy FURON, Laurent AMSALEG and Yannis AVRITHIS2.1 Introduction 412.1.1 A bit of history and vocabulary 422.1.2 Machine learning 442.1.3 The classification of images by deep neural networks 462.1.4 Deep Dreams 482.2 Adversarial images: definition 492.3 Attacks: making adversarial images 512.3.1 About white box 522.3.2 Black or gray box 622.4 Defenses 642.4.1 Reactive defenses 642.4.2 Proactive defenses 662.4.3 Obfuscation technique 672.4.4 Defenses: conclusion 682.5 Conclusion 682.6 References 69Chapter 3 Codes and Watermarks 77Pascal LEFEVRE, Philippe CARRE and Philippe GABORIT3.1 Introduction 773.2 Study framework: robust watermarking 783.3 Index modulation 813.3.1 LQIM: insertion 813.3.2 LQIM: detection 823.4 Error-correcting codes approach 823.4.1 Generalities 843.4.2 Codes by concatenation 863.4.3 Hamming codes 883.4.4 BCH codes 903.4.5 RS codes 933.5 Contradictory objectives of watermarking: the impact of codes 963.6 Latest developments in the use of correction codes for watermarking 983.7 Illustration of the influence of the type of code, according to the attacks 1023.7.1 JPEG compression 1033.7.2 Additive Gaussian noise 1063.7.3 Saturation 1063.8 Using the rank metric 1083.8.1 Rank metric correcting codes 1093.8.2 Code by rank metric: a robust watermarking method for image cropping 1133.9 Conclusion 1213.10 References 121Chapter 4 Invisibility 129Pascal LEFEVRE, Philippe CARRE and David ALLEYSSON4.1 Introduction 1294.2 Color watermarking: an approach history? 1314.2.1 Vector quantization in the RGB space 1324.2.2 Choosing a color direction 1334.3 Quaternionic context for watermarking color images 1354.3.1 Quaternions and color images 1354.3.2 Quaternionic Fourier transforms 1374.4 Psychovisual approach to color watermarking 1394.4.1 Neurogeometry and perception 1394.4.2 Photoreceptor model and trichromatic vision 1414.4.3 Model approximation 1444.4.4 Parameters of the model 1454.4.5 Application to watermarking color images 1464.4.6 Conversions 1474.4.7 Psychovisual algorithm for color images 1484.4.8 Experimental validation of the psychovisual approach for color watermarking 1514.5 Conclusion 1554.6 References 157Chapter 5 Steganography: Embedding Data Into Multimedia Content 161Patrick BAS, Remi COGRANNE and Marc CHAUMONT5.1 Introduction and theoretical foundations 1625.2 Fundamental principles 1635.2.1 Maximization of the size of the embedded message 1635.2.2 Message encoding 1655.2.3 Detectability minimization 1665.3 Digital image steganography: basic methods 1685.3.1 LSB substitution and matching 1685.3.2 Adaptive embedding methods 1695.4 Advanced principles in steganography 1725.4.1 Synchronization of modifications 1735.4.2 Batch steganography 1755.4.3 Steganography of color images 1775.4.4 Use of side information 1785.4.5 Steganography mimicking a statistical model 1805.4.6 Adversarial steganography 1825.5 Conclusion 1865.6 References 186Chapter 6 Traitor Tracing 189Teddy FURON6.1 Introduction 1896.1.1 The contribution of the cryptography community 1906.1.2 Multimedia content 1916.1.3 Error probabilities 1926.1.4 Collusion strategy 1926.2 The original Tardos code 1946.2.1 Constructing the code 1956.2.2 The collusion strategy and its impact on the pirated series 1956.2.3 Accusation with a simple decoder 1976.2.4 Study of the Tardos code-Skori´c original 1996.2.5 Advantages 2026.2.6 The problems 2046.3 Tardos and his successors 2056.3.1 Length of the code 2056.3.2 Other criteria 2056.3.3 Extensions 2076.4 Research of better score functions 2086.4.1 The optimal score function 2086.4.2 The theory of the compound communication channel 2096.4.3 Adaptive score functions 2116.4.4 Comparison 2136.5 How to find a better threshold 2136.6 Conclusion 2156.7 References 216Chapter 7 3D Watermarking 219Sebastien BEUGNON, Vincent ITIER and William PUECH7.1 Introduction 2207.2 Preliminaries 2217.2.1 Digital watermarking 2217.2.2 3D objects 2227.3 Synchronization 2247.3.1 Traversal scheduling 2247.3.2 Patch scheduling 2247.3.3 Scheduling based on graphs 2257.4 3D data hiding 2307.4.1 Transformed domains 2317.4.2 Spatial domain 2317.4.3 Other domains 2327.5 Presentation of a high-capacity data hiding method 2337.5.1 Embedding of the message 2347.5.2 Causality issue 2357.6 Improvements 2367.6.1 Error-correcting codes 2367.6.2 Statistical arithmetic coding 2367.6.3 Partitioning and acceleration structures 2377.7 Experimental results 2387.8 Trends in high-capacity 3D data hiding 2407.8.1 Steganalysis 2407.8.2 Security analysis 2417.8.3 3D printing 2427.9 Conclusion 2427.10 References 243Chapter 8 Steganalysis: Detection of Hidden Data in Multimedia Content 247Remi COGRANNE, Marc CHAUMONT and Patrick BAS8.1 Introduction, challenges and constraints 2478.1.1 The different aims of steganalysis 2488.1.2 Different methods to carry out steganalysis 2498.2 Incompatible signature detection 2508.3 Detection using statistical methods 2528.3.1 Statistical test of chi2 2528.3.2 Likelihood-ratio test 2568.3.3 LSB match detection 2618.4 Supervised learning detection 2638.4.1 Extraction of characteristics in the spatial domain 2648.4.2 Learning how to detect with features 2698.5 Detection by deep neural networks 2708.5.1 Foundation of a deep neural network 2718.5.2 The preprocessing module 2728.6 Current avenues of research 2798.6.1 The problem of Cover-Source mismatch 2798.6.2 The problem with steganalysis in real life 2798.6.3 Reliable steganalysis 2808.6.4 Steganalysis of color images 2808.6.5 Taking into account the adaptivity of steganography 2818.6.6 Grouped steganalysis (batch steganalysis) 2818.6.7 Universal steganalysis 2828.7 Conclusion 2838.8 References 283List of Authors 289Index 293

O autorze

William Puech is Professor of Computer Science at Université de Montpellier, France. His research focuses on image processing and multimedia security in particular, from its theories to its applications.