The rapid proliferation of deepfake technology has raised serious concerns about the authenticity and integrity of digital media. This study addresses the urgent need for reliable and efficient deepfake face recognition systems by proposing a cost-sensitive deep learning approach for detecting manipulated video content. The proposed model utilizes the InceptionResNetV2 convolutional neural network (CNN) architecture, chosen for its strong feature extraction capabilities and computational efficiency. To enhance detection accuracy, key frame extraction was employed to identify the most informative frames within each video, reducing redundancy and processing time. The model’s performance was evaluated using the FaceForensics++ benchmark dataset, demonstrating superior accuracy and robustness across diverse manipulation techniques. Experimental results show that the proposed system achieves over 90% accuracy in distinguishing authentic from fabricated videos, validating its adaptability and effectiveness. The findings highlight the potential of deep learning–based models to serve as a foundation for trustworthy digital media verification. Nonetheless, ongoing refinement remains essential as both deepfake generation and detection technologies continue to advance.

This project investigated the potential of InceptionResNetV2 for detecting deepfakes in videos using the FaceForensics++ dataset. We implemented a system that first preprocessed the video by extracting frames and identifying faces. Each cropped face image was then fed into the pre-trained InceptionResNet V2 model. The model assigned a probability score to each image, indicating the likelihood of it being a deepfake. Finally, by averaging the individual scores and applying a threshold, we classified the entire video as real or deepfake. This approach leveraged the ability of InceptionResNet V2 to recognize subtle inconsistencies often present in deepfaked faces. The success of this project demonstrates the potential of deep learning models for combating the spread of deepfakes.

1.W. M. Wubet, “The deepfake challenges and deepfake video detection,” Int. J. Innov. Technol. Explor. Eng, vol. 9, 2020. 2.Mitra, S. P. Mohanty, P. Corcoran, and E. Kougianos, “A machine learning based approach for deepfake detection in social media through key video frame extraction,” SN Computer Science, vol. 2, pp. 1–18, 2021. 3.M. Tanvir Rouf Shawon, G. Shahariar Shibli, F. Ahmed, and S. K Saha Joy, “Explainable cost-sensitive deep neural networks for brain tumor detection from brain mri images considering data imbalance,” arXiv e-prints, pp. arXiv–2308, 2023. 4.Z.-H. Zhou and X.-Y. Liu, “Training cost-sensitive neural networks with methods addressing the class imbalance problem,” IEEE Transactions on knowledge and data engineering, vol. 18, no. 1, pp. 63–77, 2005. 5.G. Lee and M. Kim, “Deepfake detection using the rate of change between frames based on computer vision,” Sensors, vol. 21, no. 21, p. 7367, 2021. 6.Rossler, D. Cozzolino, L. Verdoliva, C. Riess, J. Thies, and M. Nießner, “Faceforensics++: Learning to detect manipulated facial images,” in Proceedings of the IEEE/CVF international conference on computer vision, 2019, pp. 1–11. 7.Dolhansky, J. Bitton, B. Pflaum, J. Lu, R. Howes, M. Wang, and C. C. Ferrer, “The deepfake detection challenge (dfdc) dataset,” arXiv preprint arXiv:2006.07397, 2020 8.Xu, J. Liu, J. Liang, W. Lu, and Y. Zhang, “Deepfake videos detection based on texture features.” Computers, Materials & Continua, vol. 68, no. 1, 2021. 9.P. Korshunov and S. Marcel, “Deepfakes: a new threat to face recognition? assessment and detection,” arXiv preprint arXiv:1812.08685, 2018. 10.Y. Li, X. Yang, P. Sun, H. Qi, and S. Lyu, “Celeb-df (v2): a new dataset for deepfake forensics [j],” arXiv preprint arXiv, 2019. 11.Kohli and A. Gupta, “Detecting deepfake, faceswap and face2face facial forgeries using frequency cnn,” Multimedia Tools and Applications,vol. 80, pp. 18461–18478, 2021. 12.Kim, S. Tariq, and S. S. Woo, “Fretal: Generalizing deepfake detection using knowledge distillation and representation learning,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 1001–1012 13.F. Chollet, “Xception: Deep learning with depthwise separable convolu tions,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), July 2017.