From Deepfakes to Digital Truths: The Role of Watermarking in AI-Generated Image Verification
Document Type
Conference Proceeding
Publication Date
1-1-2024
Abstract
The evolution of Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) has introduced deepfake technology. Deepfake technology is a form of digital manipulation that alters video, image, and audio content with the help of Generative AI. Those deepfakes have increased concerns in various fields, including education, art, and they also raise ethical and security concerns due to their potential for deceptive content. Reviewing the increasing challenge of identifying high-quality deepfakes, there's a pressing need for robust measures to counter them. This review explores various watermarking techniques and their use to protect content authenticity and origin. Watermarking embeds a subtle watermark and provides a strong defense against deepfake technologies and similar AI-driven tools. The paper discusses current watermarking methods, their strengths and weaknesses, and potential improvements to verify AI-generated content.
Identifier
85201152152 (Scopus)
ISBN
[9798350365597]
Publication Title
2024 47th International Conference on Telecommunications and Signal Processing, TSP 2024
External Full Text Location
https://doi.org/10.1109/TSP63128.2024.10605975
First Page
216
Last Page
222
Recommended Citation
Thakkar, Jinal Jagdishkumar and Kaur, Arashdeep, "From Deepfakes to Digital Truths: The Role of Watermarking in AI-Generated Image Verification" (2024). Faculty Publications. 921.
https://digitalcommons.njit.edu/fac_pubs/921