Center for Biomedical Engineering, School of Information Science and Technology, Fudan University, Shanghai, 200433, PR China.
Center for Biomedical Engineering, School of Information Science and Technology, Fudan University, Shanghai, 200433, PR China; Key Laboratory of Medical Imaging Computing and Computer Assisted Intervention of Shanghai 200093, China.
Comput Biol Med. 2022 Aug;147:105654. doi: 10.1016/j.compbiomed.2022.105654. Epub 2022 May 21.
Photoplethysmography (PPG), as one of the most widely used physiological signals on wearable devices, with dominance for portability and accessibility, is an ideal carrier of biometric recognition for guaranteeing the security of sensitive information. However, the existing state-of-the-art methods are restricted to practical deployment since power-constrained and compute-insufficient for wearable devices. 1D convolutional neural networks (1D-CNNs) have succeeded in numerous applications on sequential signals. Still, they fall short in modeling long-range dependencies (LRD), which are extremely needed in high-security PPG-based biometric recognition. In view of these limitations, this paper conducts a comparative study of scalable end-to-end 1D-CNNs for capturing LRD and parameterizing authorized templates by enlarging the receptive fields via stacking convolution operations, non-local blocks, and attention mechanisms. Compared to a robust baseline model, seven scalable models have different impacts (-0.2%-9.9%) on the accuracy of recognition over three datasets. Experimental cases demonstrate clear-cut improvements. Scalable models achieve state-of-the-art performance with an accuracy of over 97% on VitalDB and with the best accuracy on BIDMC and PRRB datasets performing 99.5% and 99.3%, respectively. We also discuss the effects of capturing LRD in generated templates by visualizations with Gramian Angular Summation Field and Class Activation Map. This study conducts that the scalable 1D-CNNs offer a performance-excellent and complexity-feasible approach for biometric recognition using PPG.
光电容积脉搏波描记术（PPG）作为可穿戴设备上使用最广泛的生理信号之一，具有便携性和可访问性的优势，是生物识别的理想载体，可确保敏感信息的安全性。然而，现有的最先进方法受到限制，无法在实际部署中使用，因为可穿戴设备的功率和计算能力有限。一维卷积神经网络（1D-CNN）在顺序信号的众多应用中取得了成功。尽管如此，它们在建模长程依赖关系（LRD）方面存在不足，而 LRD 在基于高安全性 PPG 的生物识别中是极其需要的。鉴于这些限制，本文对可扩展的端到端 1D-CNN 进行了比较研究，以通过堆叠卷积操作、非局部块和注意力机制来扩大接收场，从而捕获 LRD 和参数化授权模板。与强大的基准模型相比，七个可扩展模型在三个数据集上对识别精度的影响（-0.2%至 9.9%）不同。实验案例表明了明显的改进。可扩展模型在 VitalDB 上的识别精度超过 97%，在 BIDMC 和 PRRB 数据集上的精度最高，分别达到 99.5%和 99.3%。我们还通过 Gramian Angular Summation Field 和 Class Activation Map 可视化来讨论捕获生成模板中 LRD 的效果。这项研究表明，可扩展的 1D-CNN 为使用 PPG 进行生物识别提供了性能卓越且复杂度可行的方法。
