| Abstract [eng] |
The PhD thesis analyses problems related with usability of human – computer interfaces based on physiological computing. The human – computer interfaces based on physiological computing are controlled by physiological signals that a human generates. These physiological signals can be distorted by various physiological states (e.g. fatigue, emotions etc.) of the human, which are changing over time. Hence, the performance of using this kind of interfaces may decrease. The adaptive human – computer interface is one of the methods to solve the problem. The model of the adaptive human – computer interface based on physiological computing, called human-assistive HCI model, is described in this work. This conceptual model describes the means of control of the specific human – computer interface as well as feedback, which is initiated by the system after the initial user state changes. The proposed model has been applied to develop 3 different systems: (1) gaze spelling system, (2) eye-controlled game and (3) EMG-based speller. Each of the aforementioned systems can react to the change of user performance, which usually decreases, when fatigue emerges. The implementation of 3 different systems based on human-assistive HCI model demonstrates the versatility of the model, also validates it. Moreover, the quantitative research of implemented human – computer interfaces is presented in this dissertation. |
