Copyright © 2026 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{180296,
        author = {Gaurav Bisht and Arnav Godhamgaonkar and Pratik Gavit and Harshil Poshiya and Prof. J. S. Pawar},
        title = {Deepfake Detection Using GAN Discriminators: Implementation and Result Analysis},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {1},
        pages = {1069-1077},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=180296},
        abstract = {Differentiating between actual and deepfakes becomes increasingly challenging in digital forensics as synthetic picture production advances. This paper offers a practical approach to detect images generated by a Vanilla Generative Adversarial Network (GAN) by use of the discriminator of another vanilla generator. Since the GAN discriminator is adversarially trained to uncover minute deviations from true data distributions, it is naturally suited for deepfake detection unlike traditional classifiers. Following Vanilla GAN training, we ran a fine-tuning step in which the discriminator acted as a stand-alone classifier. In controlled settings the model achieved a final detection accuracy of 100%. This work significantly clarifies the possibilities of adversarial components as forensic tools by offering information on the technique, architecture, training phases, and performance results. Future directions demand evaluating cross-GAN generalization and enhancement of model robustness for pragmatic environments.},
        keywords = {Deepfake Detection, Generative Adversarial Networks (GANs), GAN Discriminator, Synthetic Image Classification, Adversarial Training, Vanilla GANs, Discriminator Fine-Tuning.},
        month = {June},
        }