Comparison of Deepfake Detection Techniques through Deep Learning

Deepfakes are realistic-looking fake media generated by deep-learning algorithms that iterate through large datasets until they have learned how to solve the given problem (i.e., swap faces or objects in video and digital content). The massive generation of such content modification technologies is rapidly affecting quality public discourse safeguarding human rights. being widely used as a malicious source misinformation court seek sway court’s decision. Because evidence critical outcome many legal cases, detecting deepfake extremely important high demand forensics. As such, it identify build classifier can accurately distinguish between authentic disguised media, especially facial-recognition systems be identity protection too. In this work, we compare most common, state-of-the-art face-detection classifiers Custom CNN, VGG19, DenseNet-121 using an augmented real dataset. Data augmentation boost performance reduce computational resources. Our preliminary results indicate VGG19 has best highest accuracy 95% when compared with other analyzed models.