Advances in Deepfake Detection: A Simple Review
Abstract
Anjali Kshatriya, Tulsi Patel and Komal Prajapati
Deepfake technology, powered by AI and deep learning, provides hyper-realistic synthetic media that may serve entertainment uses as well as endanger reality with misinformation, privacy, and ethical threats. This surveyed study provides a review of recent progress in deepfake detection for facial manipulation, voice synthesis, and audio-visual forgery, detailing techniques that employ CNNs, graph neural networks, BiLSTM with attention, multimodal fusion, and meta-learning. We also discuss broader challenges of real-world deployment, dataset bias, fairness, and cross-domain generalization. We aim to provide a resource for researchers wanting to research better deepfake detection techniques and mitigate the misuse of deepfake technology in digital ecosystems.