@misc{indiciaeef1b8d594fee,
  title = {Improving the Perturbation-Based Explanation of Deepfake Detectors Through the Use of Adversarially-Generated Samples},
  author = {Konstantinos Tsigos and Evlampios Apostolidis and Vasileios Mezaris},
  year = {2025},
  url = {https://arxiv.org/abs/2502.03957},
  note = {Source identifier: 2502.03957}
}