Performance Optimization of Surface Electromyography Based Biometric Sensing System for Both Verification and Identification

Abstract

Recently, surface electromyography (sEMG) emerged as a novel biometric authentication method. EMG system parameters, such as the feature sets and channel numbers, have been known to affect system performances. Therefore, it is crucial to investigate these effects on the performance of the sEMG-based biometric system to determine optimal system parameters. In this study, three robust feature sets, Time-domain (TD) feature, Frequency Division Technique (FDT), and Autoregressive (AR) feature, and their combinations were investigated, while the number of channels varying from one to eight. For these system parameters, the performance of sixteen static wrist and hand gestures was systematically investigated in two authentication modes: verification and identification. The results from 24 participants showed that the TD features significantly (p<0.05) and consistently outperformed FDT and AR features for all channel numbers. The results also showed that the performance of a four-channel setup was not significantly different from those with higher numbers of channels. The average equal error rate (EER) for a four-channel sEMG verification system was 4% for TD features, 5.3% for FDT features, and 10% for AR features. For an identification system, the average Rank-1 error (R1E) for a four-channel configuration was 3% for TD features, 12.4% for FDT features, and 36.3% for AR features. The electrode position on the flexor carpi ulnaris (FCU) muscle had a critical contribution to the authentication performance. Thus, the combination of the TD feature set and a four-channel sEMG system with one of the electrodes positioned on the FCU are recommended for optimal authentication performance.

Synthetic EMG Based on Adversarial Style Transfer can Effectively Attack Biometric-based Personal Identification Models

“…HitRate = AttackerW antedOutputs/T otalAttacks (9) Conf usionRate = W rongOutputs/T otalAttacks (10) Wrong outputs included outputs that were what the attacker wanted and outputs that were not what the attacker wanted but were still not correct. The data are averaged across eighteen subjects, and all the values are Under attacks of the synthetic EMG signals, identification models' rank-1 and rank-5 evaluation metrics dropped close to zero, which means that these identification models were disabled.…”

“…Jiang et al [8] utilized HD-sEMG signals of common daily hand gestures as identification inputs and proposed a cancelable HD-sEMG-based biometrics system to protect personal information security. Pradhan et al [9] designed a series of experiments on the effect of different feature extraction methods and the number of channels to the EMG-based identification system and systematically investigated the performance of sixteen static wrist and hand gestures.…”

“…HitRate = AttackerW antedOutputs/T otalAttacks (9) Conf usionRate = W rongOutputs/T otalAttacks (10) Wrong outputs included outputs that were what the attacker wanted and outputs that were not what the attacker wanted but were still not correct. The data are averaged across eighteen subjects, and all the values are percentages.…”

Incremental Adaptive Gesture Classifier for Upper Limb Prostheses

“…The researchers achieve a high classification accuracy of approximately 99.1% for all force levels. Three robust feature sets are used in [11], the time domain features (TD), frequency division technique (FDT) features, and autoregressive features. The researchers show that TD features significantly outperform the FTD and the AR features even when reducing the number of channels.…”

“…The researchers show that TD features significantly outperform the FTD and the AR features even when reducing the number of channels. However, both studies [10], [11] are performed in an offline classification setup. Many researchers improved the gesture recognition by employing spatial features that are obtained from HD-sEMG signals [12]- [14].…”