A Survey on Multimodal Deepfake Detection System
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Deepfake technology has advanced rapidly, leveraging deep learning to manipulate image, audio, and video content with increasing realism.These developments pose significant threats to digital security, privacy, and trust.Traditional detection methods, which typically focus on unimodal analysis (analysing only video or audio), are becoming insufficient against sophisticated multimodal deepfakes.This survey paper reviews the current state of multimodal deepfake detection.We analyse existing literature to highlight the shift from unimodal to multimodal detection strategies, examining how the fusion of audio, video, and image analysis enhances detection accuracy.Furthermore, this survey identifies key challenges such as cross-modal synchronization, dataset scarcity, and generalization across domains.Finally, we discuss ethical considerations, including fairness and explainability, and outline future directions for robust forensic analysis
Fields of Study & Subjects
Digital Media Forensic Detection
Misinformation and Its Impacts
Advanced Steganography and Watermarking Techniques
Publication details
- Journal / source
- Advances in intelligent systems research/Advances in Intelligent Systems Research
- Pages
- 1202–1210
- Publication date
- 1 April 2026
- Language
- EN
- Document type
- Book chapter