Hiding Faces in Plain Sight: Defending DeepFakes by Disrupting Face Detection
This addresses the problem of preventing individuals from becoming victims of DeepFake videos, though it is incremental as it adapts existing adversarial attacks for a specific purpose.
The paper tackles the problem of DeepFake video generation by proposing a proactive defense framework, FacePoison, that sabotages face detection to disrupt training or synthesis of DeepFake models, with experiments showing effectiveness against eleven DeepFake models.
This paper investigates the feasibility of a proactive DeepFake defense framework, {\em FacePosion}, to prevent individuals from becoming victims of DeepFake videos by sabotaging face detection. The motivation stems from the reliance of most DeepFake methods on face detectors to automatically extract victim faces from videos for training or synthesis (testing). Once the face detectors malfunction, the extracted faces will be distorted or incorrect, subsequently disrupting the training or synthesis of the DeepFake model. To achieve this, we adapt various adversarial attacks with a dedicated design for this purpose and thoroughly analyze their feasibility. Based on FacePoison, we introduce {\em VideoFacePoison}, a strategy that propagates FacePoison across video frames rather than applying them individually to each frame. This strategy can largely reduce the computational overhead while retaining the favorable attack performance. Our method is validated on five face detectors, and extensive experiments against eleven different DeepFake models demonstrate the effectiveness of disrupting face detectors to hinder DeepFake generation.