User State Recognition Using Wavelet Transform and Singular Value Decomposition

Yuki Maruno (1051105)

We propose a novel user state recognizer for mobile applications. Since the applications should not consume too much electric power, we aim to develop the algorithm which has not only high accuracy but also low electric power consumption by using just a single three-axis accelerometer. In feature extraction with the wavelet transform, we employ the Haar mother wavelet that allows low computational complexity. In addition, we reduce dimensions of features by using the singular value decomposition. In spite of the complexity reduction, we could classify the user states into walking, running, standing still and being in a moving train with an accuracy of over 90%.