A stimulus-response based EEG biometric using mallows distance
- PDF / 2,430,225 Bytes
- 12 Pages / 595.276 x 790.866 pts Page_size
- 2 Downloads / 181 Views
REGULAR PAPER
A stimulus‑response based EEG biometric using mallows distance Baicheng Chen1 · Kun Woo Cho2 · Chenhan Xu1 · Zhengxiong Li1 · Feng Lin3 · Zhanpeng Jin1 · Wenyao Xu1 Received: 23 November 2019 / Accepted: 27 June 2020 © China Computer Federation (CCF) 2020
Abstract Electroencephalogram (EEG) activity from the brain is a promising biological marker that can serve as personal identification. Despite substantial efforts, it remains an unsolved problem to quantify EEG feature distribution for brain biometrics due to the complexity of the brain. In this study, we attempt to tackle EEG-based identification challenges by exploiting a novel distribution model. The distribution dissimilarity is measured by Mallows distance, a cluster similarity sensitive distance that is robust to signal noises. Specifically, EEG signals are decomposed through several statistical feature extraction methods, autoregressive model, discrete wavelet transform, and fast Fourier transform. With the dataset obtained from the real-world application, our proposed system achieves the f score accuracy of 96.18% and half total error rate of 2.223% , which demonstrates the feasibility and effectiveness of utilizing EEG biometrics for personal identification applications. Keywords Biometrics · Secure authentication · Wearable computing
1 Introduction Personal identification has been a long-lasting problem around the world. Due to the large population of humans, existing methods often fail with a lack of identification accuracy in system design. Researchers are actively attempting to solve this problem, yet, solutions often compromise quickly due to the computational complexity and performance speed * Baicheng Chen [email protected] Kun Woo Cho [email protected] Chenhan Xu [email protected] Zhengxiong Li [email protected] Feng Lin [email protected] Zhanpeng Jin [email protected] Wenyao Xu [email protected] 1
University at Buffalo, The State University of New York, Buffalo, NY, USA
2
Princeton University, Princeton, NJ, USA
3
Zhejiang University, Hangzhou, Zhejiang, China
trade-off. In wearable computing application scenarios, authentication methods with high computational complexity require a longer time, while rapid authentications are often quickly reverse-engineered. In the past decades, the technique of using biometrics for personal identification to secure information has gained massive popularity (Jain et al. 2008; Tian et al. 2018). Anthropocentric traits are now playing pivotal roles in everyday user authentication for both privacy and security purposes. From authenticating a user through fingerprint, face or iris, banking service authentication using voice, medical analysis using gait, to governmental agency utilizing DNA toward citizen identification (Jain et al. 1997; Beenau et al. 2006; Raja et al. 2015; Brumback et al. 2015), biometric applications cover a wide variety of scenarios. Moreover, innovative biometric modalities, including palm vessel (Li et al. 2018; Wang et al. 2018), heartbeat (Lin e
Data Loading...
