Copyright © 2025 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{187179,
        author = {Aakanksha B. Bodke and Raviraj R. Bhakare and Chaitali D. Bonde and Neha A. Chaudhari},
        title = {StyleGAN-2 ADA Signature Extraction with Recall-Enhanced CNNs for Static Deepfake Image Detection},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {6},
        pages = {3460-3465},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=187179},
        abstract = {Deepfake technology leverages advanced generative models to create hyper-realistic synthetic images that can mislead viewers and propagate misinformation. With the rapid improvement in their visual quality, distinguishing deepfakes from authentic images has become increasingly difficult. This work presents a hybrid detection framework that integrates StyleGAN2-ADA for generating realistic samples with a Convolutional Neural Network (CNN) for classification. The proposed system extracts latent image features to differentiate real from manipulated content and incorporates generative replay to mitigate catastrophic forgetting, enabling the model to retain knowledge of earlier deepfake patterns while adapting to new ones. Experimental outcomes highlight the potential of this approach to improve robustness, adaptability, and accuracy in deepfake detection},
        keywords = {Deepfake detection, Convolutional Neural Networks (CNNs), StyleGAN2-ADA, Generative replay, Catastrophic forgetting},
        month = {November},
        }