Human daily activity recognition on sequential sensor data in smart homes
With the increase of average life expectancy at birth and decrease in birth rate, the group of elderly is the fastest growing segment compared to any other age group. In comparison to younger people, the elderly are more vulnerable to experience cognitive and/or physical changes. It is clearly diffi...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2019
|
| Subjects: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | With the increase of average life expectancy at birth and decrease in birth rate, the group of elderly is the fastest growing segment compared to any other age group. In comparison to younger people, the elderly are more vulnerable to experience cognitive and/or physical changes. It is clearly difficult to rely solely on the increasing number of caregivers. Furthermore, many older people prefer to stay in their own homes as long as possible and to remain independent. In order to support the elderly, smart homes have been introduced. A smart home is a residential home settings augmented with a diversity of sensors, actuators and devices to collect information about the occupant. Smart homes support the elderly by monitoring their activities and detecting potential dangerous activities. The aim of activity recognition is to infer the activities of the occupant from a series of sensor readings. Most of the existing work in activity recognition assumes that activities have been segmented or deal with segmentation and recognitions separately. In order to apply activity recognition in the real world, both segmentation and recognition should not be treated separately. In this thesis, the issue of simultaneous segmentation and activity recognition is addressed. |
|---|