A Case Study on How Beautification Filters Can Fool Deepfake Detectors

Alexandre Libourel (1) , Sahar Husseini (1, 2) , Nelida Mirabet-Herranz (1) , Jean-Luc Dugelay (1)

1 Eurecom [Sophia Antipolis]
2 Docaposte Groupe

If telling the difference between a real video and a deepfake is difficult, with the proliferation of beautification filters on social networks it becomes nearly impossible to differentiate between a real video, a video enhanced by a filter, and a video with its original identity replaced. Therefore, is it possible to fool state-of-the-art (SotA) detectors by simply applying a beautification filter to the manipulated video? In this paper, we study the impact of beautification filters on Celeb-DF-B, a novel database created by applying popular social media beautification filters to a subset of real and fake videos from the Celeb-DF dataset. We assessed three SotA passive deepfake detectors, comparing their performance against that of human evaluators. The results indicate that filters significantly alter the behavior of the three detectors studied, resulting in a notable decrease in the video-level AUC when classifying beautified videos. In the context of human-level performance, the use of filters similarly influences human decision-making, affecting the accurate categorization of videos as either real or fake.

Mots clés: Deepfake detection; Social media filters; Beautification; Subjective evaluation

Domaines: Informatique [cs]

Alexandre Libourel, Sahar Husseini, Nelida Mirabet-Herranz, Jean-Luc Dugelay. A Case Study on How Beautification Filters Can Fool Deepfake Detectors. IWBF 2024, 12th IEEE International Workshop on Biometrics and Forensics, IEEE, Apr 2024, Twente, Netherlands. ⟨hal-04514781⟩