The photo search engine announced at the DEMO Conference on emerging technologies that the public phase of its beta test will begin in this calendar quarter.
When all of one’s photos have less-than-descriptive names and number in the thousands, finding one photo containing a particular person or people can be a chore. Unless the user went through the process of tagging each photo with descriptive metadata at some point, there was no simple way to sift through those images.
Riya will soon move its answer to the photo search problem into the public eye. A public beta test of the technology should be available in early 2006. Company CEO Munjal Shah disclosed at DEMO why he helped form Riya:
“I have 37,343 digital photos in my personal collection. Each is labeled DSC009.jpg. I can’t find anyone unless I go through the tedious process of manually labeling — or tagging — each of my photos.”
Getting Riya up to speed on one’s personal photos means training Riya to recognize people. The company noted in a statement it takes a few minutes to train Riya to identify people, but once that is done the search engine can go through a photo collection to find and tag images with those persons.
“Riya is not perfect yet, but it sure beats manually tagging each photo,” Shah told the conference.
The proprietary technology behind Riya includes some interesting features. It reads the context of a photo and considers clothing worn by people and the location of the photo. Also, Riya uses text recognition to read words appearing in photos.
At one point in 2005, a possible Google purchase of Riya in the $30 million to $40 million range began circulating. But after the company’s formal launch, those rumors passed unfulfilled.
—
document.write(“Email murdok here.”)
Drag this 
Add to document.write(“Del.icio.us”) | DiggThis | Yahoo My Web
David Utter is a staff writer for murdok covering technology and business.
