Compact Models for Periocular Verification Through Knowledge Distillation

AutorBoutros, Fadi; Damer, Naser; Fang, Meiling; Raja, Kiran; Kirchbuchner, Florian; Kuijper, Arjan
ArtConference Paper
AbstraktDespite the wide use of deep neural network for periocular verification, achieving smaller deep learning models with high performance that can be deployed on low computational powered devices remains a challenge. In term of computation cost, we present in this paper a lightweight deep learning model with only 1.1m of trainable parameters, DenseNet-20, based on DenseNet architecture. Further, we present an approach to enhance the verification performance of DenseNet-20 via knowledge distillation. With the experiments on VISPI dataset captured with two different smartphones, iPhone and Nokia, we show that introducing knowledge distillation to DenseNet-20 training phase outperforms the same model trained without knowledge distillation where the Equal Error Rate (EER) reduces from 8.36% to 4.56% EER on iPhone data, from 5.33% to 4.64% EER on Nokia data, and from 20.98% to 15.54% EER on cross-smartphone data.
KonferenzGesellschaft für Informatik, Special Interest Group on Biometrics and Electronic Signatures (BIOSIG International Conference) <19, 2020, Online>
ReferenzBrömme, Arslan (Ed.) et al.: BIOSIG 2020, 19th International Conference of the Biometrics Special Interest Group. Proceedings: 16.-18.09.2020, Fully Virtual Conference. Bonn: GI, 2020. (GI-Edition - Lecture Notes in Informatics (LNI). Proceedings P-306), pp. 291-298
SchlüsselISBN : 9783885797005