top

  Info

  • Utilizzare la checkbox di selezione a fianco di ciascun documento per attivare le funzionalità di stampa, invio email, download nei formati disponibili del (i) record.

  Info

  • Utilizzare questo link per rimuovere la selezione effettuata.

Biblioteca

MARC Lista (tabellare)
Adversarial Multimedia Forensics
Adversarial Multimedia Forensics
Autore Nowroozi Ehsan
Edizione [1st ed.]
Pubbl/distr/stampa Cham : , : Springer, , 2024
Descrizione fisica 1 online resource (298 pages)
Altri autori (Persone) KallasKassem
JolfaeiAlireza
Collana Advances in Information Security Series
ISBN 3-031-49803-8
Formato Materiale a stampa
Livello bibliografico Monografia
Lingua di pubblicazione eng
Nota di contenuto Intro -- Preface -- Acknowledgments -- Contents -- About the Editors -- Model Poisoning Attack Against Federated Learning with Adaptive Aggregation -- 1 Introduction -- 2 Related Work -- 2.1 Federated Learning in Image Classification -- 2.2 Poisoning Attacks in Federated Learning -- 2.3 Adaptive Aggregation in Federated Learning -- 3 Background and Problem Definition -- 4 Stealthy Model Poisoning Attack (MSA) -- 5 Adaptive Federated Optimization -- 6 Evaluation -- 6.1 Simulation Setup -- 6.2 Parameter and System Setting -- 6.3 Test Metrics -- 6.4 Dataset -- 7 Experimental Results -- 7.1 Results in No Attack Scenario -- 7.2 Results With Attack in Rounds Multiples of 5 -- 7.3 Results With Attack in rounds multiples of 10 -- 7.4 Results With Attack in Rounds Multiples of 20 -- 8 Discussion on the Results -- 9 Conclusions and Future Work -- References -- Image Forgery Detection Using Comprint: A Comprehensive Study -- 1 Introduction -- 2 State of the Art -- 2.1 Active Forgery Detection Methods -- 2.2 Passive Forgery Detection Methods -- 3 Comprint -- 3.1 Fingerprint Extraction -- 3.2 Forgery Localization -- 3.3 Fusing Comprint and Noiseprint -- 4 Evaluation -- 4.1 Measures -- 4.2 Artificial Dataset: Differences in JPEG Parameters -- 4.3 Realistic Datasets: Real and Manipulated Images -- 5 Limitations -- 6 Conclusion -- References -- Good or Evil: Generative Adversarial Networks in Digital Forensics -- 1 Introduction -- 2 Background -- 2.1 Forensics vs Anti Forensics vs Counter-Anti Forensics -- 2.2 Generative Adversarial Networks -- 2.3 Methodology -- 3 GAN-Based Anti-forensics -- 3.1 Image Manipulation Anti-forensics -- 3.2 Anti-forensics for Computer-Generated Images -- 3.3 GAN-Based Audio Anti-forensics -- 3.4 Advantages of GAN-Based Anti-forensics -- 4 Overview of GAN Methods to Improve the Forensic Detectors.
4.1 Strengthening Forensic Detectors to Identify Deep Fakes -- 4.2 Improving Signature Forensics Using GANs -- 4.3 Summary of GAN Usage to Improve the Detection of Forged and Synthetic Data -- 5 Future Directions -- 6 Conclusion -- References -- Refined GAN-Based Attack Against Image Splicing Detection and Localization Algorithms -- 1 Introduction -- 2 Related Work -- 2.1 Siamese Network Based Forensic Algorithms -- 2.2 Traditional Anti-Forensic Attacks -- 2.3 GAN-Based Approaches -- 2.4 Adversarial Example Attacks -- 2.5 Attacks Against Clustering and Contrastive Learning -- 2.6 Multi-Task Training Optimization -- 3 Problem Formulation -- 4 Proposed Method -- 4.1 Stage 1. Patch Sampling -- 4.2 Stage 2. Refinement -- Loss Functions -- Refinement Training -- Deployment -- 5 Experimental Setup -- 5.1 Dataset -- 5.2 Training Methods -- Construction of Communities -- Sampling from Communities -- Training Parameters -- 5.3 Evaluation Metrics -- 6 Results -- 6.1 Overall Attack Performance -- 6.2 When Baseline Generative Attack Fails -- Attack Against Splicing Detection -- Attack Against Splicing Localization -- 6.3 Ablation Study -- Impact of LDn -- Impact of LSn -- Impact of Multi-Optimizer -- 7 Limitations -- 7.1 Improvement from Good Baseline Generative Attacked Images -- 7.2 Uncontrollable Localization Heatmap -- 7.3 Optimal Loss Weights and Learning Rates -- 8 Conclusion -- References -- Generative Adversarial Networks for Artificial Satellite Image Creation and Manipulation -- 1 Introduction -- 2 Relevant Work -- 3 Datasets -- 4 DNN Models -- 4.1 Pix2Pix for Season Transfer -- Architecture -- Loss Function -- Training -- 4.2 CycleGAN for Land Cover Transfer -- CycleGAN Architecture -- Loss Function -- Training -- 4.3 Splicing wiht iGPT -- Architecture -- 4.4 Splicing Using Vision Transformer -- Architecture -- Training -- 5 Results -- 6 Conclusion.
References -- Domain Specific Information Based Learning for Facial ImageForensics -- 1 Introduction -- 2 Multimedia Forensics -- 3 General Framework of Automated Forensic Face Recognition -- 4 Related Work -- 5 Challenges of Face Recognition for Forensics -- 6 Motivation -- 7 Datasets Used -- 7.1 Composite Sketch with Age Variations Dataset -- 7.2 E- PRIP Dataset -- 7.3 Disguised Faces in the Wild Dataset -- 8 Metrics Used to Evaluate the Performance -- 9 Applications -- 10 Conclusion -- References -- Linguistic Steganography and Linguistic Steganalysis -- 1 Introduction -- 2 Preliminaries -- 2.1 Mathematical Description of Linguistic Steganography -- 2.2 General Framework of Linguistic Steganalysis -- 2.3 Language Model -- 2.4 Evaluation Metrics -- 3 Linguistic Steganography -- 3.1 Modification-Oriented Linguistic Steganography (ModLS) -- 3.2 Generation-Oriented Linguistic Steganography (GenLS) -- 4 Linguistic Steganalysis -- 4.1 Conventional Linguistic Steganalysis -- 4.2 End-to-End Linguistic Steganalysis -- 5 Challenges and Opportunities -- 6 Conclusion -- References -- Random Deep Feature Selection's Efficiency in Securing Image Manipulation Detectors Opposed by Adversarial Attacks -- 1 Introduction -- 2 Proposed Method -- 2.1 RDFS Detection Considering FC Networks -- 2.2 RDFS Detection Considering SVM -- 3 Different Implementations of Manipulation Detection -- 3.1 CNN Original Detectors -- 3.2 Minimizing Detectors -- 3.3 Adversarial Attacks -- 4 Discussion and Experimental Findings -- 4.1 Experimental Design -- 4.2 Classification Based on FC Results -- 4.3 Classification Based on SVM Results -- 5 Conclusion -- References -- Optimized Distribution for Robust Watermarking of Deep Neural Networks Through Fixed Embedding Weights -- 1 Introduction -- 2 DNN Watermarking History and Prior Art -- 3 Notation, Watermarking Model and Requirements.
3.1 Preliminaries -- 3.2 Watermark Specifications and Model -- 4 Suggested DDN Watermarking Method -- 4.1 Reliable Watermarking Algorithm -- Watermark Embedding -- Optimizing the Distribution of Weights with Water Markers -- 4.2 Watermark Extraction -- 5 Experimental Methodology -- 5.1 Host Networks and Tasks -- 5.2 Watermarking Algorithm Configuration -- 5.3 Setting for Robustness Analysis -- 6 Findings and Analysis -- 6.1 Performance of Watermarked DNN Models -- Analysis of Weights Distribution (Watermark Security) -- 6.2 Robustness Assessment -- Model Compression -- Retraining -- Comparison with the State of the Art -- 7 Conclusion Stage -- References -- Anti Forensic Measures and Their Impact on Forensic Investigations -- 1 Introduction -- 2 Factors of AFM Effectiveness -- 2.1 Problems Regarding the Effectiveness -- 2.2 Discussed Factors -- 3 Anti Forensic Measures -- 3.1 Physical Destruction -- 3.2 Artifact Wiping -- 3.3 Disk Wiping -- 3.4 NSRL Scrubbing -- 3.5 MACE Scrambling -- 3.6 File Signature Scrambling -- 3.7 Registry Wiping -- 3.8 Log Manipulation -- 3.9 Defragmentation -- 3.10 Artifact Injection -- 3.11 File Compression -- 3.12 Forensic Tool Bug Exploitation -- 3.13 Compression Bombs -- 3.14 Loop References -- 3.15 Steganography -- 3.16 Data Hiding Techniques -- 3.17 Live Operating Systems -- 3.18 Virtual Machines -- 3.19 Non-standard Programs -- 3.20 Data Pooling -- 3.21 Non-standard RAIDs -- 3.22 Dummy Drives -- 3.23 Full Disk/Drive, and File Encryption -- 3.24 Encrypted Containers -- 3.25 File Encryption -- 3.26 Hash Collisions -- 3.27 Media Integrity Obfuscation -- 4 Conclusion -- References -- Using Vocoder Artifacts For Audio Deepfakes Detection -- 1 Introduction -- 2 Related Works on Audio Deepfake -- 2.1 Human Voice Synthesis -- 2.2 Neural Vocoders -- 2.3 Autoregressive Models -- 2.4 Diffusion Based Models.
2.5 GAN Based Models -- 2.6 Synthetic Audio Detection -- 3 Vocoder Detection -- 3.1 Speaker Verification Based Rawnet2 -- 3.2 Vocoder Identification Based Rawnet2 -- 4 The LibriVoc Datasets -- 4.1 Librivoc High-Level Overview -- 4.2 Training Process -- 4.3 Train/Dev/Test Split Configuration -- 5 Evaluation Results -- 5.1 Synthesized Audio Detection -- 5.2 Classification of Neural Vocoder Artifacts -- 5.3 Leave N Out Cross-Validation -- 5.4 Robustness and Ablation Studies -- 6 Conclusion -- References -- Index.
Record Nr. UNINA-9910842298403321
Nowroozi Ehsan  
Cham : , : Springer, , 2024
Materiale a stampa
Lo trovi qui: Univ. Federico II
Opac: Controlla la disponibilità qui
IInformation fusion in distributed sensor networks with Byzantines / / Andrea Abrardo [et al.]
IInformation fusion in distributed sensor networks with Byzantines / / Andrea Abrardo [et al.]
Autore Abrardo Andrea
Edizione [1st edition 2021.]
Pubbl/distr/stampa Singapore : , : Springer Singapore : , : Imprint : Springer, , 2021
Descrizione fisica 1 online resource (XIX, 109 p. 26 illus., 15 illus. in color.)
Disciplina 006.25
Collana Signals and Communication Technology
Soggetto topico Signal processing
Image processing
Speech processing systems
Computer security
System safety
ISBN 981-329-001-3
Formato Materiale a stampa
Livello bibliografico Monografia
Lingua di pubblicazione eng
Nota di contenuto Introduction -- Preliminaries -- Security attacks and defenses in distributed sensor networks -- A Heuristic defense: A soft isolation algorithm for adversarial decision fusion -- Optimum decision fusion in the presence of Byzantines -- Nearly optimum decision fusion via message passing -- Decision fusion of hidden-Markov observations with synchronized attacks -- Decision fusion with unbalanced priors -- Decision fusion game with incomplete knowledge in a Bayesian setup -- Conclusions.
Record Nr. UNINA-9910484226603321
Abrardo Andrea  
Singapore : , : Springer Singapore : , : Imprint : Springer, , 2021
Materiale a stampa
Lo trovi qui: Univ. Federico II
Opac: Controlla la disponibilità qui