We present an experimental study of automatic camera control in the performance of collaborative remote repair tasks using video-mediated communication. Twelve pairs of participants, one "helper" and one "worker," completed a series of Lego puzzle tasks using both a static camera and an automatic camera system that was guided in part by tracking the worker's hand position. Results show substantial performance benefits for the automatic system, particularly for complex tasks. The implications of these results are discussed, along with some lessons for the use of motion tracking as a driver for camera control. Author Keywords Camera control, computer-supported cooperative work, collaboration, video mediated communication, video conferencing, motion tracking, empirical studies. ACM Classification Keywords H.5.3 Group and Organization Interfaces ? Computersupported Cooperative Work
Abhishek Ranjan, Jeremy P. Birnholtz, Ravin Balakr