Knock Knock: A Binary Human–Machine Interactive Channel for Smartphone with Accelerometer

A binary human–machine interactive channel for smartphone with accelerometer is proposed in this paper. By comparing and analyzing the advantages and disadvantages of the four binary gestures, the knock gesture is selected as the interactive gesture. Subsequently, we elaborate on the knock gesture-oriented binary human–computer interaction channel. The accelerometer signal is sampled during the interaction process. Three methods including the heuristic algorithm, the support vector machine algorithm, the online sliding window and bottom-up algorithm are used to cut the sampled data into bit signal segment. Three machine learning algorithms including the decision tree, the support vector machine, and the naive Bayesian are separately adopted to transform the cut signal segment into bit information. Finally, our binary human–computer interaction channel is verified by experiments. A higher recognition rate can be achieved only by using traditional machine learning methods.