Convolutional Neural Network-Based Detection of Facial Image Forgeries Using Binary Classification

Ahmed Samaila1; Etemi. J. Garba2; Corresponding author

doi:10.56201/ijcsmt.vol.12.no9.2026.pg95.104

📄 Download Paper

Convolutional Neural Network-Based Detection of Facial Image Forgeries Using Binary Classification

Ahmed Samaila1 and Etemi. J. Garba2, Corresponding author

Abstract

Automated forgery detection has become a significant issue in image-forensics due to the increasing realism of digitally altered or artificially created facial photographs. In order to classify facial photos as either real or fake, a convolutional neural network was created and put through experimental testing. Stratified sampling was used to reduce the collection to 2,041 face photos. During training, images were scaled to 224 by 224 pixels, normalized, and enhanced. The model was assessed using accuracy, precision, recall, F1-score, confusion- matrix analysis, and receiver operating characteristic-area under the curve (ROC-AUC) after being trained using Google Colab's Adam optimization method. On the provided test data, the suggested CNN obtained an accuracy of 59.02%, precision of 56.67%, recall of 53.13%, F1- score of 54.84%, and ROC-AUC of 0.6303. The confusion matrix reported 70 true negatives, 39 false positives, 45 false negatives and 51 true positives. Training and validation behavior showed a large generalization gap, which is consistent with over-fitting. The comparative results for MesoNet-4, MobileNetV2 and Xception showed higher classification performance than the proposed CNN, but the proposed model had a smaller number of parameters than Xception and less reported processing time. Our results suggest that the model has learned some discriminative information related to facial forgeries, but further generalization, wider dataset diversity and validation are needed before practical forensic deployment.

Keywords

facial image forgery; deepfake detection; convolutional neural network; image forensics; binary classification

References

Afchar, D., Nozick, V., Yamagishi, J., & Echizen, I. (2018). MesoNet: A compact facial video forgery detection network. 2018 IEEE International Workshop on Information Forensics and Security , 1–7. https://doi.org/10.1109/WIFS.2018.8630761 Chesney, R., & Citron, D. K. (2019). Deep fakes: A looming challenge for privacy, democracy, and national security. California Law Review, 107(6), 1753–1820. LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521, 436–444. https://doi.org/10.1038/nature14539 Li, Y., Yang, X., Sun, P., Qi, H., & Lyu, S. (2020). Celeb-DF: A new dataset for deepfake forensics. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 3207–3216. Mirsky, Y., & Lee, W. (2021). The creation and detection of deepfakes: A survey. ACM Computing Surveys, 54(1), Article 7. https://doi.org/10.1145/3425780 Rössler, A., Cozzolino, D., Verdoliva, L., Riess, C., Thies, J., & Nießner, M. (2019). FaceForensics++: Learning to detect manipulated facial images. Proceedings of the IEEE/CVF International Conference on Computer Vision, 1–11. Tolosana, R., Romero-Tapiador, S., Vera-Rodriguez, R., Fierrez, J., & Ortega-Garcia, J. (2020). DeepFakes and beyond: A survey of face manipulation and fake detection. Information Fusion, 64, 131–148. https://doi.org/10.1016/j.inffus.2020.06.014 Verdoliva, L. (2020). Media forensics and deepfakes: An overview. IEEE Journal of Selected Topics in Signal Processing, 14(5), 910–932. https://doi.org/10.1109/JSTSP.2020.3002101 Wang, S.-Y., Wang, O., Zhang, R., Owens, A., & Efros, A. A. (2020). CNN-generated images are surprisingly easy to spot... for now. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 8695–8704.

Convolutional Neural Network-Based Detection of Facial Image Forgeries Using Binary Classification

Abstract

Keywords

References

More Articles from INTERNATIONAL JOURNAL OF COMPUTER SCIENCE AND MATHEMATICAL THEORY