A Multi-layer Feature Fusion Model for Biometric Systems
Sorin Soviany, Sorin Puşcoci, Virginia Cristiana Sandulescu, Florin Şerbănescu · 2021 International Conference on e-Health and Bioengineering (EHB) · 2021
This paper presents a data fusion model for the multimodal biometric systems design. The combination of the extracted features is applied within the same biometric modality (intra-modal feature fusion) and between different biometrics (inter-modal feature fusion). The model is used with a hierarchical classifier which is trained to accurately recognize a certain biometric identification. This is a more challenging process than the verification because the one-to-many matching (multi-class classification) has a higher complexity degree than the one-to-one matching (binary classification) as it involves a larger searching space or more than 2 classes to be matched. The proposed data fusion method can be used within a home teleassistance persons based on behavioral pattern framework to provide the informative support for proper decisions.