The goal of this project was to study to what extent can mid-air gesture-based interaction be used for interacting with a PC.
We developed a mid-air gesture-based interface for basic web browser interaction based on the functionality of a Kinect sensor.
The interface was implemented incrementally based on development and evaluation of three prototypes.
The interface supports the ability to control the cursor, clicking elements, scrolling up and down and navigating forward and backward in the web history using gestures.
Furthermore we modified the original Keystroke-Level Model for prediction of task completion times using the mid-air gesture-based interface and validated the model through an experiment.
Lastly we conducted a field study on how users approached, learned and experienced the use of the mid-air gesture-based interface in public.