Representation Learning for EEG-Based Biometrics Using Hilbert–Huang Transform

• Published in: Computers (Basel) Vol. 11; no. 3; p. 47
• Main Authors: Svetlakov, Mikhail, Kovalev, Ilya, Konev, Anton, Kostyuchenko, Evgeny, Mitsel, Artur
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01.03.2022
• Subjects: Accuracy; Artificial intelligence; Biometrics; Classification; Computer architecture; Datasets; Discriminant analysis; EEG; Electrodes; Electroencephalography; Facial recognition technology; Hilbert–Huang transform; Learning; Machine learning; multi-similarity loss; Neural networks; Parkinson's disease; Recording sessions; representation learning; Signal processing; Spectrograms; subject-independent; Time series
• ISSN: 2073-431X; 2073-431X
• DOI: 10.3390/computers11030047

A promising approach to overcome the various shortcomings of password systems is the use of biometric authentication, in particular the use of electroencephalogram (EEG) data. In this paper, we propose a subject-independent learning method for EEG-based biometrics using Hilbert spectrograms of the data. The proposed neural network architecture treats the spectrogram as a collection of one-dimensional series and applies one-dimensional dilated convolutions over them, and a multi-similarity loss was used as the loss function for subject-independent learning. The architecture was tested on the publicly available PhysioNet EEG Motor Movement/Imagery Dataset (PEEGMIMDB) with a 14.63% Equal Error Rate (EER) achieved. The proposed approach’s main advantages are subject independence and suitability for interpretation via created spectrograms and the integrated gradients method.