User performance of alternative input device in human-computer interaction

Abstract

Good user interface design works from the user````s capabilities and limitations, not the computer````s. This applies to generic interfaces for larger groups of people as well as to designing special interfaces for disabled user. The main aim of this study was to develop a method for measuring and evaluating user performance using an eye-controlled input device and a three-dimensional (3D) input device. 3D tracking technique is a key factor for the eye-controlled input device and the 3D device. We developed a 3D tracking system based on passive vision technique. This system has dynamic accuracy with 0.7 % average error and 4.2 % maximum error, and the sampling rates to 6 - 10 Hz. In this study, a series of three experiments were conducted to investigate human performance variations using an eye-controlled input device and a series of two experiments were conducted to investigate those using a 3D input device. In experiment 1, the D/C gain of a mouse was optimized. The mouse was used to compare with the eye-controlled input device for the following experiments. In experiment 2, eight subjects were studied using the mouse and the eye-controlled input device. Target selection methods were clicking for the mouse and performing an intentional blink for the eye-controlled input device. The experiment investigated movement times and selection times in discrete movement acquisition tasks. The indices of performance were 10.8 bits/s for the mouse and 6.2 bits/s for the eye-controlled input device. In experiment 3, the experimental task was identical to that of experiment 2 except the direction of the movement was added as an independent variable. The effect of direction on movement time was not statistically significant. The indices of performance were 10.6 bits/s for the mouse and 6.7 bits/s for the eye-controlled input device. The results of the three experiments indicated that Fitts paradigm would be useful for evaluation an eye-controlled input device. In exp...