Proceedings of the 9th ACM Conference on Embedded Networked Sensor Systems
Proceedings of the 2012 ACM Conference on Ubiquitous Computing
Reconfigurable natural interaction in smart environments: approach and prototype implementation
Personal and Ubiquitous Computing
| Hi-index | 0.00 |
Interaction with mobile devices that are intended for everyday use ischallenging since such systems are continuously optimized towards smalloutlines. Watches are a particularly critical as display size, processingcapabilities, and weight are tightly constraint.This work presents a watch device with an integrated gesture recognitioninterface. We report the resource-optimized implementation of our algorithmicsolution on the watch and demonstrate that the recognition approach isfeasible for such constraint devoices.The system is wearable during everydayactivities and was evaluated with eight users to complete questionnairesthrough intuitive one-hand movements.We developed a procedure to spot and classify input gestures from continuousacceleration data acquired by the watch. The recognition procedure is based onhidden Markov models~(HMM) and was fully implemented on a watch.Thealgorithm achieved an average recall of 79\% at 93\% precision in recognizingthe relevant gestures. The watch implementation of continuous gesture spottingshowed a delay below 3\,ms for feature computation, Viterbi path processing,and final classification at less than 4\,KB memory usage.