Mouse and Keyboard Cursor Warping to Accelerate and Reduce the Effort of Routine HCI Input Tasks

David Rozado demonstrates how eye tracking can be used to warp the mouse cursor to the user's point of regard, accelerating input tasks.

Gaze tracking has traditionally been suggested as a possible alternative to traditional pointing mechanisms for computer input. However, the accuracy limitations of gaze estimation algorithms has constrained the potential of gaze tracking to become a reliable substitute of the mouse or the keyboard. Additionally, overloading a perceptual channel such as vision with a motor control task can be cumbersome and fatiguing. This work explores how to use gaze tracking to aid traditional cursor positioning methods with both the mouse and the keyboard during standard human computer interaction (HCI) tasks. The proposed approach consists of eliminating a large portion of the manual effort involved in cursor movement by warping the cursor to the estimated point of regard (PoR) of the user on the screen as estimated by video-oculography gaze tracking. With the proposed approach, bringing the mouse cursor or the keyboard cursor to a target position on the screen still involves a manual task but the effort involved is substantially reduced in terms of mouse movement amplitude or number of keystrokes performed. This is accomplished by the cursor warping from its original position on the screen to whatever position the user is looking at when a single keystroke or a slight mouse movement is detected. The user adjust then the final fine-grained positioning of the cursor manually. Requiring the user to be looking at the target position to bring the cursor there only requires marginal adaptation on the part of the user since most of the time, that is the default behavior during standard HCI. This work has carried out an extensive user study on the effects of cursor warping in common computer tasks involving cursor repositioning. The results show how cursor warping using gaze tracking information can speed up and reduced the physical effort required to complete several common computer tasks: mouse/trackpad target acquisition, text cursor positioning, mouse/trackpad/keyboard based text selection and drag and drop operations. The effects of gaze tracking and cursor warping on some of these tasks have never been studied before. The results show unequivocally that cursor warping using gaze tracking data can significantly speed up and reduce the manual effort involved in HCI for most, but not all, of the previously listed tasks.