Track 5： Frontiers in Multimedia Forgery Detection and Anti-Forgery Technologies: Challenges, Innovations, and Applications | 多媒体伪造检测与反伪造技术前沿：挑战、创新与应用

Organizers

Ke Xu

Chair

Associate Professor
School of Computer Science, Shanghai Jiao Tong University

Qiang Xu

Co-Chair

Assistant Professor
School of Computer Science, Shanghai Jiao Tong University

Yuezun Li

Co-Chair

Assistant Professor / Lecturer
Faculty of Information Science and Engineering, Ocean University of China

Abstract

Frontiers in Multimedia Forgery Detection and Anti-Forgery Technologies: Challenges, Innovations, and Applications

With the rapid advancement of generative AI, deep learning, and multimedia editing tools, the creation of realistic forged multimedia content (e.g., deepfakes, manipulated images/videos, synthetic audio) has become increasingly accessible. While these technologies bring convenience to creative industries, they also pose severe threats to information authenticity, personal privacy, social trust, and even national security. This academic forum aims to serve as an interdisciplinary platform for researchers, engineers, policymakers, and practitioners worldwide to exchange cutting-edge insights, innovative methodologies, and practical experiences in multimedia forgery and detection.​

We will focus on the latest progress in forgery generation techniques, the development of robust detection algorithms, the construction of high-quality datasets, and the exploration of regulatory and ethical frameworks. By bridging the gaps between academia and industry, theory and practice, this forum seeks to promote collaborative research, address critical challenges such as cross-modal forgery, adversarial attacks, and real-time detection, and contribute to the development of reliable, scalable solutions for combating misinformation and protecting digital authenticity. We welcome original research papers, review articles, and case studies that advance the state-of-the-art in this dynamic and vital field.​

 

Topics

  • Multimedia Forgery Generation Techniques
  • Deepfake creation (face swapping, voice synthesis, video reenactment)
  • Image and video manipulation (inpainting, splicing, content-aware editing)
  • Synthetic multimedia content (AI-generated text-to-image/video/audio)
  • Cross-modal forgery technologies (e.g., text-to-deepfake, audio-driven video forgery)
  • Detection and Forensics Methodologies
  • Deep learning-based detection algorithms (CNN, Transformer, GAN-based detectors)
  • Digital forensic techniques (metadata analysis, noise pattern recognition, sensor fingerprinting)
  • Cross-domain and generalized detection for unseen forgery types
  • Real-time and low-latency detection systems for edge devices
  • Datasets, Evaluation, and Benchmarks
  • Large-scale and diverse datasets for forgery detection and forensics
  • Standardized evaluation metrics and benchmarking frameworks
  • Comparative analysis of state-of-the-art methods
  • Challenges in dataset construction (privacy protection and realism)
  • Challenges and Emerging Issues
  • Adversarial attacks and defense strategies for detection systems
  • Forgery detection in low-quality or compressed multimedia
  • Ethical, legal, and social implications (privacy, accountability, regulation)
  • Applications in critical domains (journalism, law enforcement, healthcare, social media)

Invited Speakers(more to be added)

  • Laijin Meng
    Shanghai Jiao Tong University
  • Zhongjie Mi
    City University of Hong Kong