Presentation | 2008-03-23 User Context Recognition for Use in Wearable Learning Systems Based on Congestion Level Estimation of the Inside of a Train Using a Carbon Dioxide Sensor Tomonori NAKAMURA, Takefumi OGAWA, Kiyoshi KIYOKAWA, Haruo TAKEMURA, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | In this paper, we have investigated a user context recognition mechanism for a wearable system that provides a context-based uninterrupted learning environment based on congestion level estimation of the inside of a train using a carbon dioxide sensor. In our research, we measure the acceleration of a user's legs, ceiling height, and carbon dioxide level in the area to recognize the following five user conditions: sitting, standing, walking, running, and biking, the following three standing locations: on a train, on a platform, and at a crossing gate or a traffic signal, and the following two congestion levels of the inside of a train: uncrowded and crowded. The support vector machine predicts the probability of a user's context based on the power spectrum of the acceleration data, the median of the height data, and the raw carbon dioxide data. In addition, we introduce context transition tendency to achieve robust recognition of user context. We have conducted a series of experiments to evaluate our prototype. Our system was able to recognize user context during an actual commute with an accuracy of 85.8%. Also, we confirmed that the misrecognition rate that results in an interference with learning where the system gives a question the user cannot answer is extremely low (0.6%). |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Wearable Learning System / Context Recognition / Accelerometer / Ultrasonic Sensor / Carbon Dioxide Sensor |
Paper # | MVE2007-89 |
Date of Issue |
Conference Information | |
Committee | MVE |
---|---|
Conference Date | 2008/3/15(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Media Experience and Virtual Environment (MVE) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | User Context Recognition for Use in Wearable Learning Systems Based on Congestion Level Estimation of the Inside of a Train Using a Carbon Dioxide Sensor |
Sub Title (in English) | |
Keyword(1) | Wearable Learning System |
Keyword(2) | Context Recognition |
Keyword(3) | Accelerometer |
Keyword(4) | Ultrasonic Sensor |
Keyword(5) | Carbon Dioxide Sensor |
1st Author's Name | Tomonori NAKAMURA |
1st Author's Affiliation | Graduate School of Information Science and Technology, Osaka University() |
2nd Author's Name | Takefumi OGAWA |
2nd Author's Affiliation | Information Technology Center, The University of Tokyo |
3rd Author's Name | Kiyoshi KIYOKAWA |
3rd Author's Affiliation | Graduate School of Information Science and Technology, Osaka University:Cybermedia Center, Osaka University |
4th Author's Name | Haruo TAKEMURA |
4th Author's Affiliation | Graduate School of Information Science and Technology, Osaka University:Cybermedia Center, Osaka University |
Date | 2008-03-23 |
Paper # | MVE2007-89 |
Volume (vol) | vol.107 |
Number (no) | 554 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |