A new system developed by Rice University researchers could enable mobile users to quickly determine their location indoors without having to communicate with the cloud, networks, or other devices.
The technology uses image recognition and "hashing," a method that reduces key details in a photo to short strings of numbers. Camera-based Positioning System Using Learning (CaPSuLe) hashes a photo from the user's camera and compares it against a pre-downloaded, highly compressed location database, called a hash table, to determine his or her location. In tests using a commercially available smartphone, CaPSuLe calculated locations in less than two seconds with greater than 92% accuracy using less than 4 joules of energy.
The proof-of-concept application uses a combination of machine learning and inexact computing to address privacy, computations, and energy challenges. The core hashing-based image-matching algorithm is more than 500 times cheaper--both in terms of energy and computational overhead--than state-of-the-art image-matching techniques.
"CaPSuLe shows that a 'cloudless,' probabilistic approach can be a viable and more sustainable alternative" than cloud-based machine-learning applications, says CaPSule co-inventor and Rice professor Anshumali Shrivastava.
From Rice University
View Full Article
Abstracts Copyright © 2016 Information Inc., Bethesda, Maryland, USA
No entries found