The MMUISD Gait Database and Performance Evaluation Compared to Public Inertial Sensor Gait Databases


Permatasari, Jessica and Tee, Connie and Ong, Thian Song (2020) The MMUISD Gait Database and Performance Evaluation Compared to Public Inertial Sensor Gait Databases. In: Computational Science and Technology. Lecture Notes in Electrical Engineering (Computational Science and Technology), 603 . Springer Verlag, pp. 189-198. ISBN 9789811500572

[img] Text
10.1007@978-981-15-0058-9.pdf - Published Version
Restricted to Repository staff only

Download (46MB)


Gait recognition is an emerging biometric method that allows an automatic verification of a person by the way he or she walks. This paper presents a new dataset for gait recognition using mobile sensors called MMUISD Gait Database that resembles the real world as closely as possible. The existing public gait databases are acquired in controlled settings. In this study, an Android application is developed to record the human gait signals through inertial measurement unit sensor such as accelerometer and gyroscope with 50 Hz fixed sampling rate. A preliminary evaluation with 80 samples of participant’s data is carried out to assess the gait recognition performance using the new dataset. Time and frequency domain are used to extract gait features from the raw sensors data. The accuracy is assessed using eight classifiers with 10-fold cross validation. The results show that phone positions and orientation affect the gait recognition performance. The MMUISD dataset that introduces such variability provides a good opportunity for researchers to further investigate these challenges.

Item Type: Book Section
Uncontrolled Keywords: Biometric identification, Inertial Sensor, Gait Recognition, Gait Dataset
Subjects: T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7800-8360 Electronics
Divisions: Faculty of Information Science and Technology (FIST)
Depositing User: Ms Suzilawati Abu Samah
Date Deposited: 16 Dec 2020 12:11
Last Modified: 16 Dec 2020 12:11


Downloads per month over past year

View ItemEdit (login required)