• Thippareddy Karthik Reddy
• Kollabathula Rakshan Dev
• Gonugunta Chamikar Venkata Srikar
• Keesara Poojitha
• Prof. E. Srinivasa Reddy

Social media has seen widespread adop on across diverse domains such as educa on, entertainment, science, and adver sing. However, its misuse has spawned the crea on of fake images and videos through deep learning technologies, presen ng a formidable challenge in detec on. These fabricated videos, commonly known as deepfakes, are digitally manipulated footage, including altered clips of celebri es, cra ed with advanced image edi ng tools, rendering them indis nguishable to the human eye. Tradi onal so ware and technical methods struggle to effec vely iden fy such decep ve content, as creators con nuously refine their methods. The prolifera on of varied fake videos poses a societal threat. In response, we've devised a technique leveraging the potent Xcep onNet architecture, renowned for its excellence in image classifica on tasks, to spot deepfakes. Our model targets altered faces within realis c, intricate videos by tapping into the hierarchical features learned by convolu onal neural networks (CNN). Trained on a diverse dataset of 400 videos featuring an array of facial expressions, ligh ng condi ons, and se ngs, our approach ensures resilience against sophis cated deepfake methods. By extrac ng frames from the videos and categorizing them as "real" or "fake," we bolster the model's adaptability and tackle overfi ng through a suite of data augmenta on and regulariza on techniques, resul ng in an impressive 93.5% accuracy. Furthermore, the integra on of adversarial training empowers our model to discern between genuine and synthe c videos more adeptly, thus amplifying its efficacy in deepfake detec on.

Xcep onNet, CNN, Data Augmenta on, Regulariza on Techniques, Adversarial Training

"DEEP FAKE DETECTION USING MACHINE LEARNING ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 6, page no.b776-b783, June-2024, Available :http://www.jetir.org/papers/JETIR2406184.pdf

"DEEP FAKE DETECTION USING MACHINE LEARNING ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 6, page no. ppb776-b783, June-2024, Available at : http://www.jetir.org/papers/JETIR2406184.pdf