Open AccessCharacter Recognition on Time Series Data collected from Smartphone Sensors
Author(s) -
Deep Raval,
Jaymin Suhagiya,
Sukriti Macker
Publication year - 2021
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1099/1/012014
Subject(s) - accelerometer , gyroscope , computer science , series (stratigraphy) , character (mathematics) , field (mathematics) , focus (optics) , time series , real time computing , artificial intelligence , pattern recognition (psychology) , machine learning , mathematics , engineering , paleontology , physics , geometry , optics , pure mathematics , biology , aerospace engineering , operating system
In the modern era, smartphones have become part and parcel of life. The exponential increase in the number of smartphone users has given rise to a copious amount of data. Although the data produced by smartphones are of various types, this paper mainly focuses on the usage of sensory data produced by smartphones. This paper explores the field of recognizing characters with the aid of time-series sensor data. The primary focus of the research is to utilize recurrent neural networks to predict the digits 0 - 9 and characters A - Z . The objective achieved was by the help of sensor data that included the readings of Accelerometer, Magnetometer, Gyroscope, and Linear Accelerometer sensors providing information with respect to three-axis x , y , and z , having an interval of 0.01 seconds between two corresponding values. We succeeded in achieving the accuracy of 93.60% on the training data and 89.51% on the testing data.