Swedish Startup Offering Image-Based Search - InformationWeek

InformationWeek is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them.Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Software // Enterprise Applications

Swedish Startup Offering Image-Based Search

The service from Polar Rose can identify whether a photo contains images of people, find other photos of the same people, and help with identifying the subjects of photos.

Swedish startup Polar Rose says it can search images on the Web to find pictures of people, delivering better results than today's text-based search engines.

Polar Rose, which has received $5.1 million in funding from Nordic Venture Partners, plans to make its technology available next year as a plug-in for Internet Explorer and Firefox Web browsers. The company also intends to offer online social networks, photo-sharing sites, and other large Web sites royalty-free application programming interfaces to embed the Polar Rose service.

Image search on the Web is largely dependent on looking for keywords within tags or text surrounding a photo. Polar Rose claims that its combination of technology and user intelligence enables it to find people within images without having to depend on text indicators.

When pointed to an image, the Polar Rose service converts the 2D photo into a 3D model that's used to create a facial "footprint" of the image in the company's index. The 3D conversion greatly improves the matching ability by helping to eliminate pose and light in order to build generic attributes of faces.

The human side of the equation is necessary to match the footprint with an actual person. To do that, users add tags to the photo. Once this is done, Polar Rose can find images of a person based on the footprint, and is not dependent on metadata attached to a photo.

Mikkel Thagaard, vice president of business development for Polar Rose, said Wednesday the company's recognition engine would be useful in large community sites, where thousands of photos of users and friends are stored. It also could be used on photo-sharing sites where people create larger online photo albums.

"Looking inside the photo can get better results," Thagaard said. As the company's index grows, it could also be helpful in finding pictures on the public Web.

The company hopes to make money through contextual advertising that would accompany search results, Thagaard said. Polar Rose plans to announce the first Internet company to embed its service by late January 2007, and release the browser plug-ins at the end of the first quarter.

A mobile search service is expected to launch in beta in the second quarter, and its advertising program is set to rollout in the summer. By the fall of next year, the company expects to have a completed index of online photos.

We welcome your comments on this topic on our social media channels, or [contact us directly] with questions about the site.
Comment  | 
Print  | 
More Insights
10 Ways to Transition Traditional IT Talent to Cloud Talent
Lisa Morgan, Freelance Writer,  11/23/2020
What Comes Next for the COVID-19 Computing Consortium
Joao-Pierre S. Ruth, Senior Writer,  11/24/2020
Top 10 Data and Analytics Trends for 2021
Jessica Davis, Senior Editor, Enterprise Apps,  11/13/2020
White Papers
Register for InformationWeek Newsletters
Current Issue
Why Chatbots Are So Popular Right Now
In this IT Trend Report, you will learn more about why chatbots are gaining traction within businesses, particularly while a pandemic is impacting the world.
Flash Poll