Feasibility of Wrist-Worn, Cancelable, Real-Time Biometric Authentication via HD-sEMG and Dynamic Gestures

In recent years, high-density surface electromyography (HD-sEMG) has been applied for cancelable authentication of wearable devices used in healthcare applications. However, previous studies focused on the forearm or hand, whereas the wrist is more comfortable for practical use. Covering a larger area with fewer electrodes (16 channels on the wrist versus 256 channels on the forearm) also reduces the chance of electrode corruption and improves the stability of the authentication system. In addition, fast dynamic gestures are more desirable than static gestures for real-time authentication. Shorter gesture execution time also makes dynamic gesture codes harder to record by imposters. We propose a cancelable real-time biometric system via wrist-worn HD-sEMG and dynamic gestures. A wearable wristband was designed to collect HD-sEMG signals. Based on a dataset with 12 dynamic gestures, we realized a fast real-time authentication. It achieved an equal error rate (EER) of 0.0039 even when imposters entered a correct gesture code. The cancelability was confirmed with an EER of 0.0018. The real-time authentication with an authentication accuracy of 96.7% and a low latency of 64 ms makes it practical in healthcare applications. These results demonstrate the feasibility of wrist-worn biometrics, which allows secure healthcare applications without compromising the user experience.