Personal Identification using Gait Data on Slipper-device with Accelerometer

In this paper we presented a method of gait identification by slippers with an accelerometer to perform privacy-friendly personal identification. The gait data from accelerometer during walking is adopted from developed slipper devices as the personal unique data used for identification. Gait data is processed by Fast Fourier Transform to extract the frequency features and the Support vector machine (SVM) is used to identify the subject. Through assessing the different segmentation window size and various sensor positions, the results showed that an average accuracy was 95.0% using six sensors, and an average accuracy of 93.3% using three sensors placed at optimal positions.