Addressing the Challenge of Deepfake Media: Robust Detection for Ensuring Authenticity
Kanishka Juneja, N. Lalithamani · 2024
As fake content gets more sophisticated, there’s a competition between making better fakes and better detectors. In an era where seeing is no longer believing, we must sharpen our commitment to authenticity in the face of technological deception. The study looks closely at how deepfake detection model works, including the different ways researchers have previously described and methods used to detect misleading videos and images that are spread very easily nowadays through various platforms. This research centers on a comparative analysis of prominent predefined models, namely CNN, VGG16, and Xception networks, in the realm of deepfake detection. Employing these models individually and by exploring different combinations of them, our methodology aims to unravel the distinctive strengths of each architecture. By evaluating their performance comprehensively, we seek to identify accuracy, precision and robustness of deepfake detection, making advancements in the ongoing process of safeguarding digital authenticity. By doing so, our research seeks to bridge the gap between theoretical model strengths and practical application, ensuring that the identified methods align with the demands of the digital world.