Full Record

New Search | Similar Records

Author
Title Geometry Aware Compressive Analysis of Human Activities : Application in a Smart Phone Platform
URL
Publication Date
Degree MS
Discipline/Department Electrical Engineering
Degree Level masters
University/Publisher Arizona State University
Abstract Continuous monitoring of sensor data from smart phones to identify human activities and gestures, puts a heavy load on the smart phone's power consumption. In this research study, the non-Euclidean geometry of the rich sensor data obtained from the user's smart phone is utilized to perform compressive analysis and efficient classification of human activities by employing machine learning techniques. We are interested in the generalization of classical tools for signal approximation to newer spaces, such as rotation data, which is best studied in a non-Euclidean setting, and its application to activity analysis. Attributing to the non-linear nature of the rotation data space, which involve a heavy overload on the smart phone's processor and memory as opposed to feature extraction on the Euclidean space, indexing and compaction of the acquired sensor data is performed prior to feature extraction, to reduce CPU overhead and thereby increase the lifetime of the battery with a little loss in recognition accuracy of the activities. The sensor data represented as unit quaternions, is a more intrinsic representation of the orientation of smart phone compared to Euler angles (which suffers from Gimbal lock problem) or the computationally intensive rotation matrices. Classification algorithms are employed to classify these manifold sequences in the non-Euclidean space. By performing customized indexing (using K-means algorithm) of the evolved manifold sequences before feature extraction, considerable energy savings is achieved in terms of smart phone's battery life.
Subjects/Keywords Electrical engineering; CPU Usage; Human Activity Recognition; Non Euclidean Geometry; Symbolic Representation; Unit Quaternions; Unit sphere- S-3 manifold
Country of Publication us
Record ID oai:arizonastate:25156
Repository asu
Date Retrieved
Date Indexed 2020-05-01
Grantor Arizona State University

Sample Search Hits | Sample Images

…applications using minimum distance classifier and clustered KNN consume nearly the same amount of CPU resources. On the other hand naive-Bayes algorithm has a considerably higher amount of CPU usage. For benchmarking the CPU usage, studies are carried out by…

…thus increased CPU Usage (due to frequent PAA operation in SAX). This is clearly depicted by comparing the window sizes 5 and 30 in Figure 3.12a However, on the other hand, a further higher window size will have bad effects on the recognition…

…22 3.9 Symbolic Approximation: Symbols a, b, c are assigned to the piecewise aggregated time series. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3.10 Effect on SAX on CPU Computational Time…

…parameters on CPU Computational Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 4.1 Axis angle Representation…

…37 5.1 Android App . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 5.2 Battery Usage: PAA Window Size variation . . . . . . . . . . . . . . . . . . . . . . . . . 40 5.3 Battery Usage

…43 5.6 Gestures: Motion Trace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 5.7 Effects of varying Approximation parameters on Battery Usage . . . . . . . . 46 5.8 Pairwise Distance plots: Rotation…

…drastic consumption of battery power and significant utilization of the CPU resources. To solve the above problem, symbolic representations of the sensor data are considered to reduce the size of the acquired raw sampled data and subsequent feature…

…increased if an event is deemed unmissable by the GMM classifier. This approach serves to conserve the CPU resources of the smartphone. Vega et al [5] worked towards unconstrained mobile sensor-based human activity recognition where the activities…

.