A data augmentation based deep learning approach for deepfake image detection

Abstract

Deepfake technology, pushed by advanced deep learning algorithms, poses a serious threat to the integrity of visual content, potentially leading to misinformation, propaganda, and fraudulent evidence fabrication. our research proposes a rigorous framework for real and deepfake picture recognition. The suggested approach merges a transformer-based model, notably the Vision Transformer (ViT), coupled with fine-tuned Convolutional Neural Networks (CNNs).