AI-Generated Image and Video Synthesis

Deep Learning Models, Applications, and Ethical Implications in Visual Media Creation
1. Auflage 2026
ISBN: 978-1-394-40311-0
Verlag: John Wiley & Sons Inc

Buch, Englisch, 192 Seiten

Technical depth and ethical frameworks for AI visual media synthesis

Generative AI models for visual media are transforming virtual reality and biomedical imaging while raising urgent questions about deepfakes and misinformation. AI-Generated Image and Video Synthesis addresses both dimensions. A team of researchers provide algorithmic foundations alongside detection strategies, authentication methods, and regulatory analysis.

Coverage spans text-to-image generation, image-to-image translation, video synthesis, neural rendering, and 3D-aware generation. The book examines AI applications in CT, MRI synthetic data augmentation, and virtual staining for biomedical contexts. Case studies explore AI-assisted filmmaking, music videos, and style transfer. A dedicated chapter forecasts emerging trends including diffusion-transformer hybrids and autonomous generative agents.

Readers will also find: - Comparative analyses of generative models including GANs, diffusion models, and transformers with implementation guidance and code repositories for hands-on experimentation
- Deepfake detection strategies and digital content authentication techniques addressing misinformation, intellectual property rights, and emerging regulatory frameworks worldwide
- Industry case studies demonstrating real-world deployments in creative industries, surveillance systems, education, and cultural preservation applications
- Biomedical imaging applications covering synthetic data generation for CT and MRI, virtual staining techniques, and data augmentation strategies
- Practical toolkits supporting implementation and evaluation of AI synthesis techniques across professional and academic contexts

Designed for AI researchers, computer vision engineers, and graduate students studying deep learning and image processing, this book connects theoretical principles with practical deployment. The combination of technical depth, application coverage, and ethical analysis makes it a comprehensive resource for professionals navigating AI-generated visual media.

Autoren/Hrsg.

Mewada, Arvind

Ansari, Mohd. Aquib

Ahmad, Shahnawaz

Singh, Nagendra

Fachgebiete

• Technische Wissenschaften Elektronik | Nachrichtentechnik Nachrichten- und Kommunikationstechnik
• Mathematik | Informatik EDV | Informatik Informatik Künstliche Intelligenz
• Technische Wissenschaften Technik Allgemein Technische Zuverlässigkeit, Sicherheitstechnik
• Mathematik | Informatik EDV | Informatik EDV & Informatik Allgemein

Inhaltsverzeichnis

Contributors xi
Foreword xiii
About the Editors xv
Preface xvii
Acknowledgements xix
Acronyms xxi
Introduction xxiii

1 Introduction to AI-generated Image and Video Synthesis 1
Arvind Mewada, Mohd. Aquib Ansari, Shahnawaz Ahmad, and Nagendra Singh

1.1 Introduction 1
1.2 Foundations of AI-generated Media 2
1.3 Image Synthesis Techniques 5
1.4 Video Synthesis and Manipulation 9
1.5 Applications for AI-generated Media 12
1.6 Ethical and Societal Considerations 14
1.7 Future Directions and Challenges 17
1.8 Conclusion 19

2 LoomNet: An Assam Handloom Fabric Dataset 23
Anindita Das and Aniruddha Deka

2.1 Introduction 23
2.2 Methodology 25
2.3 Discussion and Future Work 30
2.4 Conclusion 31

3 Sensors-to-synthesis: Edge AI and IoT for Generative Visual Systems 35
Swati Vishnoi, Ankur Sisodia, Mayank Deep Khare, and Ajay Kumar Yadav

3.1 Introduction 35
3.2 Background and Literature Review 36
3.3 Architectural Framework 38
3.4 Methodological Framework and Workflow 39
3.5 Sensors-to-synthesis Workflow of Generative Visual Systems 40
3.6 Generative Models and Edge AI for Visual Synthesis 40
3.7 Applications of Edge-AI-driven Generative Visual Systems 42
3.8 Challenges and future directions 43
3.9 Conclusion 44

4 Detecting AI-generated Images in the Social Media Era: A Deep Learning Approach with GenReal Dataset 47
Akhil Sibi, Deepika Pantola, and Madhuri Gupta

4.1 Introduction 47
4.2 Literature Review 48
4.3 Methodology 49
4.4 Results 51
4.5 Conclusion and Future Scope 54

5 Raindrop Removal in Images and Videos Using Generative AI: A Survey 59
Mohd. Aquib Ansari, Vijay Dhote, Sonulal, Shahnawaz Ahmad, and Jiyaul Mustafa

5.1 Introduction 59
5.2 Background and Preliminaries 60
5.3 Generative AI Approaches for Raindrop Removal 62
5.4 Datasets and Evaluation Metrics 64
5.5 Applications 66
5.6 Challenges and Open Issues 68
5.7 Future Directions 69
5.8 Conclusion 70

6 A Transfer Learning Baseline and a GAN-augmentation Perspective for MRI-based Alzheimer's Disease Detection 73
Subiya Zaidi, Arvind Mewada, and Kapil Juneja

6.1 Introduction 73
6.2 Related Work 75
6.3 Materials and Methods 77
6.4 Results 80
6.5 Discussion 81
6.6 Conclusion 84

7 Advanced Foundations and Future Trends in Generative AI for Visual Media 89
Shwetang Dubey, Greetta Pinheiro, Reetu Singh, Mohd. Aquib Ansari, and Lalit Kumar

7.1 Context and Advanced Foundations 89
7.2 Technology Landscape and Mathematical Formulations for Visual Synthesis 90
7.3 Model Trajectories and Scaling Strategies for Visual Synthesis 94
7.4 Evaluation Protocols, Benchmarks, Robustness, and Alignment for Visual Media 96
7.5 Systems Efficiency, Economics, and Deployment for Visual Synthesis 98
7.6 Applications and Translational Pathways for Visual Media 101
7.7 Open Problems and Research Agenda for Visual Synthesis 103
7.8 Conclusion and Outlook for Visual Synthesis 105

8 High-resolution GAN Augmentation with Ensemble CNN Models for Accurate Skin Cancer Detection 109
Amit D. Joshi, Ananya Patil, Niruppreet Kour, Parth Vora, and Tamizharasan P. S

8.1 Introduction 109
8.2 Literature Review 110
8.3 Proposed Methodology 112
8.4 Experimental Setup 114
8.5 Results and Discussion 115
8.6 Conclusion and Future Scope 118

9 Content-aware Convolutional VAE for Anime Face Synthesis 123
Lalit Kumar, Bhupchand Kumar, Shahnawaz Ahmad, Mohd. Aquib Ansari, and Greetta Pinheiro

9.1 Introduction 123
9.2 Related Work 126
9.3 Proposed Model: Content-aware CNN-VAE 128
9.4 Experimental Setup 130
9.5 Results and Analysis 131
9.6 Conclusion 133

10 GEN-HAR: Generative Diffusion Learning for Human Activity Recognition 135
Roshni Singh, Ataus Samad, Abhilasha Sharma, Vandana Bhatia, and Abdul Aleem

10.1 Introduction 135
10.2 Related Work 136
10.3 Proposed Method: GEN-HAR 137
10.4 Experimental Analysis 138
10.5 Conclusion 142

11 Hybrid Neural Networks for Robust Deepfake Detection: Integrating CNN-RNN and Residual Attention Architectures 147
Rupesh Kumar Dewang, Arvind Mewada, Jamvant Omkar, Ayshwarya Jaiswal, and Nagendra Singh

11.1 Introduction 147
11.2 Related Work 148
11.3 Problem Statement 149
11.4 Proposed Work 150
11.5 Experiments and Results 153
11.6 Conclusion and Future Work 160

References 161
Index 163

Über Autor(innen)

Arvind Mewada, PhD, is an Assistant Professor in the School of Computer Science Engineering and Technology at Bennett University, India. His research spans natural language processing, machine learning, and deep learning, with publications in Multimedia Tools and Applications and The Journal of Supercomputing.

Mohd. Aquib Ansari, PhD, is an Assistant Professor at Galgotias University, India. A UGC-NET qualified scholar and M.Tech. Gold Medalist, his research focuses on computer vision, image processing, and human-computer interaction, with advances in surveillance systems and gesture recognition.

Shahnawaz Ahmad, PhD, is an Assistant Professor at Bennett University, India. His expertise includes cloud computing security and machine learning. He reviews for IEEE Access, Elsevier, Springer, and Wiley, and is the author of Cloud Computing: An Industrial Approach.

Nagendra Singh, PhD, is Principal of Trinity College of Engineering and Technology in India. He has published over 42 international journal articles, 9 conference papers, 3 Indian patents, and 4 books, contributing actively to IEEE and Scopus-indexed publications.