Autonomous Detection and Evaluation of Deepfakes: A Comprehensive Study

In recent times, there has been a notable advancement in deepfake techniques and the accessibility of extensive, cost-free databases. Consequently, even folks lacking technological expertise can now edit or produce visually authentic samples for various goals, both benign and harmful in nature. This paper provides a comprehensive analysis of the classification of methods used for deepfake generation and detection. The paper considers numerous factors, including the identified forgery, methodology or techniques used, evaluation metrics used for performance analysis, and the utilized dataset. By studying the development of deepfakes and the most up-to-date deepfake detecting methods, this study gives a full picture of deepfake techniques and makes it easier to come up with new, more reliable methods to fight the growing difficulty of deepfakes.