@article{monteiro2024performance,
  title = {Performance Analysis on Deep Fake Detection},
  author = {Stéphane MONTEIRO and Cristina WANZELLER and Filipe CALDEIRA},
  year = 2024,
  url = {https://ibimapublishing.com/articles/CIBIMA/2024/457767/},
  journal = {Communications of the IBIMA},
  volume = 2024,
  pages = 8,
  doi = 10.5171/2024.457767,
  abstract = {A deepfake is a type of synthetic media, image, video or audio of a person in which their physiognomy has been digitally altered, using artificial intelligence, particularly deep learning techniques, so that they appear to be someone else, typically used maliciously or to spread false information. In this study, our main goal is to thoroughly assess the effectiveness of deepfake detection algorithms by using various key performance metrics such as accuracy, precision, recall, and F1-score. The primary focus of our analysis revolves around their capability to distinguish between genuine and manipulated videos.Furthermore, our research involves a detailed examination of specific types of deepfake manipulations with the aim of identifying differences in detection accuracy and performance across these categories. We go beyond just analyzing the algorithms and investigate how characteristics of the dataset, like diversity and size, impact the detection performance of the tested algorithms.We anticipate that the results of this research will make substantial contributions to the advancement of deepfake detection technology. Furthermore, the insights obtained from this study will not only assist in refining existing detection algorithms but also offer valuable guidance for future research in the field of deepfake detection, ultimately contributing to the ongoing fight against the spread of deceptive digital media.},
  keywords = {Deepfakes, Machine Learning, Deep learning, DFDC, XCeption, ResNet, VGG},
  note = Article ID: 457767
}