Name
Abstract | ||
|---|---|---|
We propose a new activity recognition system for the daily activity by using a generative/discriminative hybrid model that can learn an activity classification model with small quantities of training data by sharing training data among different activity classes. Many existing activity recognition studies employ a supervised machine learning approach and thus require an end user's labeled training data, this approach places a large burden on the user. In this study, we assume that a user wears sensors (accelerometers) on several parts of the body such as the wrist, waist, and thigh, and by sharing sensor data obtained from only selected accelerometers (e.g., only waist and thigh sensors) among two different activity classes based on a sensor data similarity measure, the quantities of training data can be increased. For further reduction of the burden on the user, we also adopt semi-supervised approach to train the classifier in our study. |
Year | DOI | Venue |
|---|---|---|
2013 | 10.1145/2494091.2495991 | UbiComp (Adjunct Publication) |
Keywords | Field | DocType |
training data,daily activity,end user,existing activity recognition study,different activity class,sensor data similarity measure,sensor data,semi-supervised approach,new activity recognition system,activity classification model,activity recognition,semi supervised learning | Training set,Semi-supervised learning,Activity recognition,Similarity measure,End user,Accelerometer,Computer science,Artificial intelligence,Classifier (linguistics),Discriminative model,Machine learning | Conference |
Citations | PageRank | References |
3 | 0.44 | 9 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Quan Kong | 1 | 11 | 4.04 |
| Takuya Maekawa | 2 | 326 | 49.93 |