Boffins zoom in on digital image search
By
Robert Jaques
17 October 2007 07:21AM
Tags:
boffins | zoom | digital | image | search
Searching
for digital photographs is set to become easier thanks to smart
software developed by researchers at Penn State University..
The
US boffins have developed a software system that automatically tags
images as they are uploaded and improves the tags by "learning" from
user interaction with the system.
"Tagging itself is challenging
as it involves converting an image's pixels to descriptive words," said
James Wang, lead researcher and associate professor of information
sciences and technology at Penn State.
"But what is novel with
the Tagging over Time technology is that the system adapts as people's
preferences for images and words change."
The system can
accommodate evolving vocabulary and interpretations to images that
people have uploaded and are uploading to systems such as Yahoo's
Flickr.
This allows the system's vocabulary to grow, replacing old tags with more relevant and more specific new tags, Wang explained.
In
tests, the Tagging over Time technology correctly annotated four out of
every 10 images, a significant improvement over the researchers'
earlier annotation system known as Automatic Linguistic Indexing of
Pictures-Real Time.
In the previous system, pixel content of images was analysed to suggest annotations.
In
the new software, researchers have added a machine-learning component
that enables the computer to learn from user interaction with
photo-sharing systems.
"The bottom line is that the system
makes it easier to find photographs and is able to improve its
performance by itself as time passes," added Ritendra Datta, a graduate
student in computer science working with Wang.
"The advancement means time savings for consumers as well as improved searching and referral capabilities."
Copyright © 2007 vnunet.com