Context-Aware Activity Classification with Adaptive Transmission
Date Issued
2015
Date
2015
Author(s)
Lin, Chia-Chih
Abstract
With the growth of innovative wearable and mobile devices, smart applications in daily life become more complicated. Most of these applications offload all data from wearable and mobile devices to remote servers to overcome the limitations of device resources. However, offloading all the data, especially multimedia contents, requires a large number of network resources and may result in the dissatisfaction of users who use such applications. To alleviate the problem, we propose a practical system architecture which includes an adaptive transmission mechanism to reduce the network bandwidth usage. We design and implement a multimedia application, which generates a diary-like daily activity summarization, with the proposed system architecture to verify the feasibility. In the experiment with four participants wearing the wearable camera for fourteen days, the results show that over 89% of the overall bandwidth usage can be reduced with sacrificing 11% of the server-side performance via the proposed adaptive transmission mechanism.
Subjects
Wearable device
Wearable camera
Mobile app
Adaptive transmission
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-104-R01942038-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):67f9f5b8d87e27639fd224ff8edff299