PPT Slide
Using each single modality, has certainly limitations in both security and robustness.
Decision-level fusion can enhance the performance of each single modality.
Bayesian conciliation theory and AND/OR/median rules that fuse audio- and video-based authentication modalities can be used.
Fuzzy K-means, fuzzy vector quantization and median radial basis function neural networks are also proved efficient decision fusion managers.
Multi-modal authentication