Researchers from the University of Toronto have designed a new algorithm that will change the way we find photos on social media sites like Facebook and Flickr. The search tool understands relationships between individuals and uses tag locations to find your photos even if you are not tagged in them.
Social networks are dominating the digital lives of millions of people around the world. Services like Facebook can bring distant relatives and friends closer. Since photos speak and describe what people are going through better, most users choose to share pictures on social media.
There are over a trillion photos from billions of social media users out there. For instance, Facebook alone hosts almost half a trillion photos from more than a billion users. With such vast data-base, it gets difficult to find the photo you are looking for, unless you are tagged in one of them. Rather than losing certain photos, researchers from the University of Toronto have found a reasonable solution, more like a new search tool.
Parham Aarabi, a professor in The Edward S. Rogers Sr. Department of Electrical & Computer Engineering, and his former Master's student Ron Appel, have developed a new algorithm called relational social image search, which can act as a smart photo search tool. It uses the tag locations to technically understand the relationship between individuals and finally find you in untagged photos.
"If you want to search a trillion photos, normally that takes at least a trillion operations. It's based on the number of photos you have," Aarabi, said in a press release. "Facebook has almost half a trillion photos, but a billion users-it's almost a 500 order of magnitude difference. Our algorithm is simply based on the number of tags, not on the number of photos, which makes it more efficient to search than standard approaches."
Arabi and Appel tested the algorithm and surprisingly found a new way to generate maps. They tagged photos of a few buildings around the University of Toronto and ran search through other untagged campus photos.
"The result we got was of almost a pseudo-map of the campus from all these photos we had taken, which was very interesting," says Aarabi.
The latest technology does not use complicated tech algorithms like facial recognition software, to find photos. The project is currently in a test phase and once it is incorporated in to a fully fledged feature, it will not change the way people use the Facebook search but will only display more efficient results.
The project will be presented at the IEEE International Symposium on Multimedia Tuesday, December 10.
