Detecting Deepfakes: using CNN to Identify Manipulated Visual Media
N. Sunanda, K. Shailaja, Jampani Satish Babu, K Vamsi Krishna, N. LakshmanPratap, Ravi Kanth Motupalli · 2024
Deepfake detection is a rapidly evolving field with significant implications for the integrity of visual media. This review explores techniques, challenges, and future directions in deepfake detection. Traditional image forensics techniques, deep learning models, and multi-modal fusion approaches are employed to differentiate real content from deepfakes. Addressing the challenges of evolving deepfake algorithms and the need for real-time detection, future research should focus on novel architectures, training strategies, temporal and contextual information, and ethical considerations. By advancing deepfake detection methods, researchers can contribute to combating the proliferation of synthetic content and preserving the authenticity of visual media in the digital age.