Dartmouth College researchers led by Emiliano Miluzzo say that sharing data gathered by cell phone sensors could improve the accuracy of the data and the applications that rely on the data.
The team has developed Darwin, software that enables phones to recognize an event, such as the voice of its user, and how it could change due to different conditions, such as a noisy room. Darwin-enabled phones would be able to share the information with other phones that have the software.
In an experiment, the team installed the software on the phones of eight people and tested a voice recognition tool while they visited a quiet meeting room, a noisy restaurant, and a busy street. The handsets used the voice recognition software on the cell phone to identify a speaker, and the phones with Darwin compared the efforts to identify the speaker in an attempt to improve accuracy. Darwin-enabled phones correctly identified the speaker with up to 90 percent accuracy, compared with just 60 percent in a control group that did not have the software on their phones. "By relying on the resources around you, you get better results," Miluzzo says. "The idea is to make the sensing process on phones scalable."
From New Scientist
View Full Article
Abstracts Copyright © 2010 Information Inc., Bethesda, Maryland, USA