Abstract:
Face anti-spoofing (FAS) protects face recognition from presentation attacks. Handcrafted binary or pixel-wise labels supervise PA detectors in FAS methods. Handcrafted labels may not be the best way to supervise PA detectors learning intrinsic spoofing cues. We propose a novel Meta-Teacher FAS (MT-FAS) method to train a meta-teacher to supervise PA detectors better than handcrafted labels. Bi-level optimization trains the meta-teacher to supervise PA detectors learning rich spoofing cues. Bi-level optimization has two key components: 2) a higher-level training that optimizes the meta-teacher’s teaching performance by minimizing the detector’s validation loss. Our meta-teacher is explicitly trained to teach the detector (student), unlike existing teacher-student models, which focus on accuracy rather than teaching. Extensive experiments on five FAS benchmarks show that with the proposed MT-FAS, the trained meta-teacher 1) supervises better than handcrafted labels and existing teacher-student models and 2) significantly improves PA detector performance.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
