Key points are not available for this paper at this time.
Surface EMG-based gestures recognition systems are helping the disable to enjoy a better life. Academic institutes and commercial companies have been developing a lot of sEMG-based prosthesis, exoskeletons and rehabilitation systems. In this area, most studies are improving recognition accuracies by focusing on signals pre-processings, features and classifiers. But, in real-life circumstances, sEMG-based recognition performances are limited by many disturbances and variability from electrodes and subjects. In this paper, key advances of sEMG-based recognition systems are firs reviewed, then several disturbances or factors that limited sEMG-based performances and their solutions are summarized up in great detail. At last, prospections for improving sEMG-based robustness in the future are proposed.
Xu et al. (Wed,) studied this question.