Filtering approaches to accelerated consensus in diffusion sensor networks

Summary Mobile phones are equipped with a rich set of sensors, such as accelerometers, magnetometers, gyroscopes, photometers, orientation sensors, and gravity sensors. These sensors can be used for human activity recognition in the ubiquitous computing domain. Most of reported studies consider acceleration signals that are collected from a known fixed device location and orientation. This paper describes how more accurate results of basic activity recognition can be achieved with transformed accelerometer data. Based on the rotation matrix (Euler Angle Conversion) derived from the orientation angles of gyroscopes and orientation sensors, we transform input signals into a reference coordinate system. The advantage of the transformation is that it allows activity classification and recognition to be carried out independent of the orientation of sensors. We consider five user activities: staying, walking, running, ascending stairs, and descending stairs, with a phone being placed in the subject's hand, or in pants pocket, or in a handbag. The results show that an overall orientation independent accuracy of 84.77% is achieved, which is a improvement of 17.26% over those classifications without input transformation. Copyright © 2014 John Wiley & Sons, Ltd.

Section: Process Of Activity Recognitionmentioning

confidence: 99%

Human activity recognition based on transformed accelerometer data from a mobile phone

Xia

Wang

2014

“…The field of adaptive filter has been steadily studied for a broad range of areas like seismic system identification, channel estimation, echo/noise cancellation, etc. [1][2][3][4][5][6][7][8][9][10][11][12][13][14]. The affine projection algorithm [15] that uses the multiple input vectors is the representative algorithm in adaptive filtering theory, because it has faster convergence rate with the colored input than the normalized least-meansquares (NLMS) algorithm.…”

Section: Introductionmentioning

confidence: 99%

A robust affine projection sign algorithm against the high power of measurement noises

Yoo

Shin

Song

et al. 2016

SUMMARYThis paper proposes a novel robust affine projection sign algorithm (R-APSA) through a modified criterion that consists of the Euclidean norm of the sum of the difference between the present weight vector and the previous weight vectors. Because the filter update equation of the proposed R-APSA is obtained from the modified criterion, it has robustness against the high power of measurement noises. Moreover, due to the characteristic inherent in the original APSA, the proposed R-APSA performs well even though impulsive noises occur. Simulation results verify that the proposed R-APSA attains smaller steady-state errors than the existing released algorithms.

“…How to capture human activity information effectively and accurately is one of the most significant and valuable issues of ubiquitous computing, especially for user‐centric mobile applications , e.g. healthcare, rehabilitation, and gaming.…”

Section: Introductionmentioning

confidence: 99%

Smartphone‐based activity recognition independent of device orientation and placement

Guo

Chen

et al. 2015

SUMMARYTraditional activity recognition methods based on accelerometers generally assume that the orientation and placement of sensing devices are fixed. However, the orientation and placement of smartphones, the most widely used sensing devices, usually cannot be controlled in daily lives. As a result, the activity recognition performance would be greatly affected. Aiming at this problem, this paper proposes a smartphone-based activity recognition method, which is independent of device orientation and placement. First, the original 3D acceleration signals are mapped into a uniform reference coordinate system. Then, we apply multidimensional motif discovery techniques for the smartphone-based activity recognition problem. Experimental results show that our method can achieve better performance (with 85.2% precision and 82.9% recall) than other methods under the condition of unfixed orientation and placement.