Thesis
Energy efficient methods for human activity recognition
Washington State University
Master of Science (MS), Washington State University
05/2020
DOI:
https://doi.org/10.7273/000004099
Handle:
https://hdl.handle.net/2376/124912
Abstract
Continuous monitoring and recognition of complex activities is valuable for understanding human behavior and providing activity-aware services. At the same time, recognizing complex activities requires both movement and location information that can quickly drain batteries on wearable devices. In this paper we introduce CPAM (Change Point Activity Monitoring), an energy-efficient strategy for recognizing and monitoring complex activities in real time. CPAM employs unsupervised change point detection to detect likely activity transition times. By selectively triggering activity recognition only during the detected change points, CPAM extends device battery life by 2.6 times while retaining the activity recognition performance of continuous sampling. We validate our approach by collecting smartwatch data and comparing the energy consumption between CPAM (triggered AR) and non-CPAM (continuous AR) cases. Results indicate that change point detection techniques can be effective for reducing the energy footprint of sensor-based mobile applications and that automated activity labels can be used to estimate values of sensors between sampling periods.
Metrics
5 File views/ downloads
25 Record Views
Details
- Title
- Energy efficient methods for human activity recognition
- Creators
- Cristian Culman
- Contributors
- Diane Joyce Cook (Advisor) - Washington State University, School of Electrical Engineering and Computer Science
- Awarding Institution
- Washington State University
- Academic Unit
- School of Electrical Engineering and Computer Science
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University
- Identifiers
- 99900890786701842
- Language
- English
- Resource Type
- Thesis