de
en
Schliessen
Detailsuche
Bibliotheken
Projekt
Impressum
Datenschutz
zum Inhalt
Detailsuche
Schnellsuche:
OK
Ergebnisliste
Titel
Titel
Inhalt
Inhalt
Seite
Seite
Im Dokument suchen
Hammerl, Sebastian: Development of a smartphone based system to support personal journaling of daily activities. 2013
Inhalt
Acknowledgements
Introduction
Research Plan for Automatic Journaling of Daily Activities
Motivation
Goals and Research Questions
Definition of Daily Activities
Definition of Daily Activities for this Project
Applications for the Planned System
Summary
State-of-the-Art Activity Recognition
Technical Background
Computing Power and Storage
Small Sensors and Devices Available
Related Work
Life-Logging
Quantified Self Community
Localization, Location, and Transportation
Physical Activity Tracking from Accelerometer Data
Medical Use of Diary Systems
Research Systems
Memex Vision by Vannevar Bush
MyLifeBits by Microsoft
SenseCam by Microsoft
Commercial Systems
Classic Diary and Digital Counterparts
Fitbit Ultra
Facebook (Timeline)
Summary
Mobile Platforms and Technology Overview
Requirement Analysis
Platform Selection
Introduction
Apple – iOS
Google – Android
RIM – BlackBerry OS
HP / Palm – webOS
Microsoft – Windows Phone 7
Other Mobile Platforms
Conclusion
Hardware Requirements and Selection of a Suitable Smartphone
Overview
Requirements
Selection
Model Specifications and Sensors
Considerations for Software / User Interfaces
User Interface
Implementation
Thoughts on System Limitations & Resource Efficiency
Accuracy of Phone Sensors
Limitations of the Phone
Resource Efficiency
Extending the System
Smart Environments
Computer / TV Usage
Summary
Development of a System to Collect Daily Activity Data
User Study Motivation and Goals
Upcoming Challenges
Experiment Design
Preliminary Questionnaire
Preliminary Questionnaire Results
Conclusive Experiment Design
Software Implementation
Experiment Software Operating Mode
User Interface
Background Data Logging
Summary
Experimental Procedure
Current Activity
Well-Being
General Experiment Questionnaire
Experiment Conditions and Specifications
Experiment Questionnaires Results
Experiment Questionnaires Conclusion
First Experiences and Lessons Learned
Analysis and Evaluation of the Collected Daily Activity Data
Packing Raw Sensor Data and Computing Features
Data Overview and Structuring Daily Activity Data
Visualization and Evaluation of Daily Activity Data
Data Visualization of One Full Day
In-Depth Evaluation of Location Data
Evaluation of Other Feature Data
Activity Category Combinations and Transitions
Conclusion
Classification of Daily Activities Using Sensor Data
Classification Methods for Multi-Label Problems
Evaluation Metrics for Multi-Label Classification
Classification Results
Refining the Classification Approach
Discussion and Conclusion
Development of the AMARAS System
Re-defined Concepts and Goals
Application Implementation Details
Application Overview
Background Service
User Interface
Feature Extraction from Raw Sensor Data
On-Device Classification of Daily Activities
Technical Details
Application Summary
Evaluation of the AMARAS Application
Research Questions for the Study
Study Conditions
Notes and Problems During the Study
Questionnaire Results of the Final User Study
Classification Results Compared to the Last Study
In-Study Performance Analysis
Possible Enhancements
Summary, Discussion, and Conclusion
Summary, Conclusion and Outlook
Summary
Conclusion
Outlook
First Study
Second Study