Eye tracking is a promising solution to alleviate the asymmetrical bandwidth problem existing in human-computer interaction (HCI), but it is still problematic because of issues like the “Midas-touch” problem, limitations in tracking accuracy, calibration errors and drift, and especially eye cursor’s instability mainly resulting from the inherent jittery motions of the eyes. In order to improve the stability of eye cursor, we have designed two novel algorithms to counteract the eyes' jittery movements so as to improve the stability of eye cursor. Our experimental results indicate that it is feasible to overcome eye jitter and benefit eye pointing by a simple algorithm without producing obvious delay in the eye cursor's movements.

The development of human performance models has an important place in respect to both theory and practice in the field of human computer interaction. However, current studies focus on the effect of system factors (e.g, target width and amplitude between the targets) on movement time. This project has sought to establish models that can accurately include human physiological and psychological information (e.g., subjective factors) into mathematical functions. Such models will promote accuracy and will be applicable in human computer interaction input devices evaluation. We have studied both linear models and nonlinear models for pointing tasks and steering tasks.

The most commonly used solution to non-alphabet text input, both in Chinese and in Japanese, is to use a Roman alphabet-based phonetic system, such as pinyin in Chinese and romaji in Japanese, as an intermediary. The user has to select the intended character from a list of homophonic choices. An intriguing user interface design issue here is how to display these homophonic multiple candidates.  The different candidate display styles may impact on the user’s experience and performance.  We have done a study of four candidate display styles, three commonly used in commercial products (“vertical,” “horizontal,” and “compact-horizontal”) and one novel (“matrix”), together with various manual selection methods (mouse, numeric key, spacebar, cursor key, numeric keypad and unrestricted input method). This project contributes to our understanding of Japanese / Chinese text input.

This study sought to identify the best combinations of modalities through usability testing. We asked the question that how users choose different interaction modes when they work on a particular application? We have conducted two experiments to compare interaction modes on a CAD system and a map system. The experiments provided information on how users adapt to each interaction mode and the ease with which they are able to use these modes.

Direct manipulation by hand is an intuitive and simple way of positioning objects in an immersive virtual environment. However, this technique is not suitable for making precise adjustments to virtual objects in an immersive environment because it is difficult to hold a hand unsupported in midair and to then release an object at a fixed point. We proposed an alternative technique using a virtual 3D gearbox widget, which enables users to adjust values precisely.
Cooperative work in a distributed immersive virtual environment,which connects virtual reality systems with immersive projection displays such as CAVE, usually requires users to put annotations in the virtual world. Simple 3D symbols or 3D icons are useful in a fixed task; however, simple symbolical annotations are insufficient for more general cooperative work because complex abstract annotations cannot be represented by them. Therefore, we need textual annotations in an immersive projection display system. We have conducted a pilot experiment to evaluate text input methods using handheld devices for immersive virtual environments.

We developed a novel usability method called "Task Break Monitoring'' (TBM) for monitoring users' physiological indices when using the computer. In this method, users take a break with their eyes closed after each interaction with the computer. During each break, electroencephalogram (EEG), especially alpha 1 waves, electrocardiogram (ECG) and galvanic skin resistance (GSR) are monitored and recorded. We believe "TBM'' to be an important innovation in human computer research and development because the after effects of computer use have an obvious bearing on recovery time, user endurance and psychological attitude to the technology in general.