Goal: Interface design where eye movements are the primary and the only input. The interface intended for a normal PC user.

Implementation: A concept of a photo browsing application was selected as an example of an application where an interface component control is done by the eye movement only. Mouse and keyboard were completely excluded from the design consideration from the very beginning of the development. Interface layout was created by understanding the imitations of the Human Visual System (constant eye movement) and eye tracking equipment (accuracy). Visual feedback was  provided for every component and every selection state. Successful picture browsing was tested by 10 people. [read more here].

Requirements: Tobii x120, Windows XP OS.

Goal: Developing new modes of eye-gaze based interaction, e.g., saccade-based interaction or pursuit based interaction providing more faster, natural ways to communicate with a computer.

Motivation: The majority of the HCI systems use fixation duration (dwell time) as a trigger for interface actions. In such interfaces, the duration of the detected fixation initiates a “click”.  Fixation-based selection necessitates data buffering and therefore introduces a delay in the system. In computer gaming the speed of “clicking” the target is of utmost importance and can be the difference between a victory or a loss. Therefore the motivation for the project is to find new  interaction algorithms were the loss of the speed of the interaction is minimized.

Project Status: Looking for students interested in the project. The students will participate in the development and the evaluation of the new interactive algorithms. Please contact Dr. Komogortsev (ok11@txstate.edu) if interested in participation.

Goal: Eye-gaze guided interaction for World of Warcraft game.

Implementation: Eye-gaze data controls mouse cursor in the World of Warcraft game. Eye position coordinates are processed by the Two State Kalman Filter (TSKF) modeling an eye as a system with translational motion represented by position, velocity and white noise acceleration. The TSKF determines the position of the mouse cursor. Mouse click is generated by a fixation with specified duration. [read more here].

Requirements: ASL eye-tracker. World of Warcraft.

Goal: Eye-gaze guided human computer interaction.

Implementation: Eye-gaze data controls mouse cursor. Eye position coordinates are averaged out within a sliding window to provide the coordinates for the mouse cursor. Mouse click is generated by a fixation with specified duration. Both sliding window size and fixation duration can be adjusted in the program’s interface.

Requirements: Tobii eye-tracker

Download: [.zip]

Developed By: Dr. Oleg Komogortsev & Sandeep M. Gowda

Goal: Eye-gaze guided human computer interaction.

Implementation: Eye-gaze data controls mouse cursor. Eye position coordinates are averaged out within a sliding window to provide the coordinates for the mouse cursor. Mouse click is generated by an eye blink. Both sliding window size and number of blinks (defines a single click action) can be adjusted in the program’s interface.

Requirements: Tobii eye-tracker

Download: [.zip]

Developed By: Dr. Oleg Komogortsev & Sandeep M. Gowda

Goal: Eye-gaze guided human computer interaction.

Implementation: Eye-gaze data controls mouse cursor when a user is not performing any mouse movement for a short period of time. Any mouse action disables mouse control. The purpose of such approach is to enjoy the accuracy of mouse input and while increasing the speed of interaction by warping mouse cursor to a location of interest. Two important parameters must be considered in this interaction scheme: 1) time interval between the last mouse action and the onset of eye-gaze-based cursor control 2) minimum saccade amplitude that allows eye-gaze-based cursor movement. In other aspects eye-gaze interaction is similar to Eye-Gaze Interaction by Dwell Time.

Requirements: Tobii eye-tracker

Download: [.zip]

Developed By: Dr. Oleg Komogortsev & Sandeep M. Gowda

Goal: Evaluate the performance of various eye-gaze interaction modalities and a mouse.

Implementation: This tool allows to quickly comparing various interaction methods when the eye tracker is involved. Three HCI approaches are selected: regular mouse interaction, Eye-Gaze Interaction by Dwell Time and Eye-Gaze Interaction by Blinking. The interface allows selecting the modality of interaction and measure the performance time of each interaction method. The user has to complete a simple task of triggering a sequence of randomly generated numbers by employing selected interaction modality. Performance time of each task is measured in seconds.

Requirements: Tobii eye-tracker

Download: [.zip]

Developed By: Dr. Oleg Komogortsev & Sandeep M. Gowda