On Disentangling Spoof Trace for Generic Face Anti-Spoofing
This work addresses the challenge of improving generalization and interpretability in face anti-spoofing systems, which is crucial for security applications, though it appears incremental in advancing existing trace-based methods.
The paper tackles the problem of generic face anti-spoofing by designing an adversarial learning framework to disentangle spoof traces from input faces, achieving superior spoof detection performance on both seen and unseen scenarios.
Prior studies show that the key to face anti-spoofing lies in the subtle image pattern, termed "spoof trace", e.g., color distortion, 3D mask edge, Moire pattern, and many others. Designing a generic anti-spoofing model to estimate those spoof traces can improve not only the generalization of the spoof detection, but also the interpretability of the model's decision. Yet, this is a challenging task due to the diversity of spoof types and the lack of ground truth in spoof traces. This work designs a novel adversarial learning framework to disentangle the spoof traces from input faces as a hierarchical combination of patterns at multiple scales. With the disentangled spoof traces, we unveil the live counterpart of the original spoof face, and further synthesize realistic new spoof faces after a proper geometric correction. Our method demonstrates superior spoof detection performance on both seen and unseen spoof scenarios while providing visually convincing estimation of spoof traces. Code is available at https://github.com/yaojieliu/ECCV20-STDN.