Subject

EMG and ECG signals has been recently identified as a viable for biometric methods solving the authentication problem [1]. In these methods the main goal is to distinguish individuals exploiting the uniqueness of the signal per individual. In clinical settings the correct positioning of the sEMG electrodes plays a critical role for obtaining reliable estimates of the EMG variables [2].

Kind of work

The goal this thesis is to use biometric methods to identify individual muscles for the purpose of localizing the sEMG electrodes. The task for the student is to use the state-of-the-art machine learning methods (typically based on deep learning) to study correlation between the electrode placement and the measured EMG signal. The student will investigate the possibility of using the correlation for predicting the actual placement (of the electrodes) and location on the body.

Framework of the Thesis

[1] https://www.tandfonline.com/doi/abs/10.3109/03091902.2015.1021429

[2] https://www.sciencedirect.com/science/article/pii/S1050641108001181?casa_token=KDykwiTT4_AAAAAA:1cz2ufk06IsrCk5RNE-koWJ1b4Hdusu39kqDKAIlfFgekQnjI9-NRRf1_DzxXx9ThrEeOY_W5g

Number of Students

1