QUEST7: A Real Life Search Agent

in Accelerometer, Cloud, Gesture, Mobile Search, Quest7, Windows Phone 7
TitleQUEST7: A Real Life Search Agent
Publication TypeConference Paper
Year of Publication2011
AuthorsPedamallu, ER, Nguyen H, Joshi S, Zheng R
Conference NameInternational Workshop on Emerging Mobile Sensing Technologies, Systems, and Applications
Conference LocationSan Francisco
Abstract

Human eyes have a limited range during each eye fixation. As a result, searching specific text in a document or an image is time consuming. In this paper, we design QUEST7, a real life search agent combining the advantage of mobile devices and cloud computing. Utilizing Optical Character Recognition (OCR) service provided on the Azure cloud, we extract texts from images captured on a mobile device. Keywords are then highlighted for human use. Experiments show that QUEST7 is significantly faster and more accurate than human eyes for dense texts. Extensive measurements are carried out to identify sources of delays in the process.With 3G accesses, network transfer latency tends to dominate. In contrast, OCR on the cloud side contributes to most of the delay with Wi-Fi accesses.

Copyright © 2011 WiSeR Group. All Rights Reserved.