Photoplethysmogram Biometric Authentication Using a 1D Siamese Network

Abstract

In the head-mounted display environment for experiencing metaverse or virtual reality, conventional input devices cannot be used, so a new type of nonintrusive and continuous biometric authentication technology is required. Since the wrist wearable device is equipped with a photoplethysmogram sensor, it is very suitable for use for nonintrusive and continuous biometric authentication purposes. In this study, we propose a one-dimensional Siamese network biometric identification model using a photoplethysmogram. To maintain the unique characteristics of each person and reduce noise in preprocessing, we adopted a multicycle averaging method without using a bandpass or low-pass filter. In addition, to verify the effectiveness of the multicycle averaging method, the number of cycles was changed and the results were compared. Genuine and impostor data were used to verify the biometric identification. We used the one-dimensional Siamese network to verify the similarity between the classes and found that the method with five overlapping cycles was the most effective. Tests were conducted on the overlapping data of five single-cycle signals and excellent identification results were observed, with an AUC score of 0.988 and an accuracy of 0.9723. Thus, the proposed biometric identification model is time-efficient and shows excellent security performance, even in devices with limited computational capabilities, such as wearable devices. Consequently, our proposed method has the following advantages compared with previous works. First, the effect of noise reduction and information preservation through multicycle averaging was experimentally verified by varying the number of photoplethysmogram cycles. Second, by analyzing authentication performance through genuine and impostor matching analysis based on a one-dimensional Siamese network, the accuracy that is not affected by the number of enrolled subjects was derived.

Keywords: biometric; deep learning; identification; lightweight; one-dimensional Siamese network; photoplethysmogram.

Figure 1

The waveform of a photoplethysmogram (PPG) signal.

Figure 2

Example of PPG data.

Figure 3

Preprocessing steps.

Figure 4

Results of detrending: (a) original signal; and (b) processing result.

Figure 5

Results of peak detection (blue line: PPG signal): (a) result of finding the foot points (orange circles) in the PPG signal without distance; (b) result of single-cycle extraction without distance; (c) result of finding the foot points (orange circles) in the PPG signal using distance; and (d) result of single-cycle extraction using distance.

Figure 6

Results of interpolation (blue dots: original signal; gray line: a line connecting the dots corresponding to the original signal; red dots: interpolation results).

Figure 7

Multicycle averaging results (Blue lines: N single cycle signals. Red line: the averaged one) (N = 5).

Figure 8

Structure of proposed Siamese network using 1D-CNN.

Figure 9

Structure of the twin network.

Figure 10

Overall structure of the model.

Figure 11

Training loss graph according to N.

Figure 12

Comparison of loss (green: N = 1; orange: N = 4; and blue: N = 5).

Figure 13

Receiver operating characteristic (ROC) curve according to N value (purple line: N = 5; red line: N = 4; green line: N = 3; orange line: N = 2; and blue line: N = 1).

Figure 14

Genuine–impostor histogram (blue bar: impostor; and orange bar: genuine).

Figure 15

FAR–FRR trend as the change in decision threshold and EER.