Name
Abstract | ||
|---|---|---|
Wireless sensor networks, with their capability to capture physical phenomena at micro-scale, have changed how we collect and analyze data from the real world. Especially, using low-power cameras we can design interesting applications that provide us with previously difficult-to-capture information from the real-world. While cameras hold privacy threats in human-residing environments, they can be actively used in natural world analyzing applications. However, the disadvantage of using cameras is that the samples themselves (e.g., images) have large sizes, forcing the system to exchange more data, and in turn, decreasing energy efficiency. Given that camera-based sensor networks are usually deployed in remote locations, the lifetime of individual devices become a major concern. In this work, we target to utilize machine-learning algorithms to characterize the context of images captured from a real-world deployments. Specifically, using images from the James Reserve bird nest deployment [1], we utilize and optimize machine learning algorithms to operate on embedded low-power, resource-limited platforms. Using both a systematic and algorithmic approach, our proposed systems architecture and algorithms hold the potential to reduce the amount of data to be transmitted by as much a eight-fold. |
Year | Venue | Field |
|---|---|---|
2016 | RTCSA | Histogram,Computer science,Server,Visual sensor network,Real-time computing,Artificial intelligence,Systems architecture,Contextual image classification,Distributed computing,Key distribution in wireless sensor networks,Efficient energy use,Wireless sensor network,Machine learning |
DocType | Citations | PageRank |
Conference | 0 | 0.34 |
References | Authors | |
1 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Jungmo Ahn | 1 | 4 | 2.92 |
| Jeongyeup Paek | 2 | 868 | 53.50 |
| JeongGil Ko | 3 | 674 | 64.60 |