Once in a while we like to get out of the office and visit with an Alt search engine.
Today we visited Daniel McClary, CEO of Mugr, just to see what they are up to.

I’ll be honest, I don’t take a lot of pictures. However, most of my friends do– party pictures, amateur portraits, social-networking snapshots. I’m a scientist at heart, and what I really like is hard problems. I believe there’s a benefit not just to solving hard problems, but in trying to solve them. Even less-than-perfect attempts at solving a challenging problem can open the door to learning. So, when presented with a challenging problem wherein even a small amount of progress might enhance the lives of my friends, it seemed a natural fit.

Image-based search is a vertical search that often achieves more attention than it does success. At Mugr, we attempt to use face recognition technology to help people do more with their photographs. There’s something truly compelling about being able to extract the information our photographs contain. The people, places, and things we’ve chosen to memorialize in pixels or silver– they’re important, and we’d like to keep track of them.

These moments in time constitute a tremendous amount of data. Width by height obviously gives us the number of raw pixels, but that’s only where the challenge begins. Skipping past the technical, the face-based image searching Mugr attempts to provide boils down to two very human questions: Is there person’s face in this picture? Who does that face belong to?

Our species is uniquely developed to answer these questions. We see faces everywhere: the clouds, the Moon, even as far off as Mars! However, developing an effective, reliable way for computers to answer either question is still an open topic in both academics and industry. Machines can perform tremendous analysis of raw variability, avoiding human cognitive peculiarities, yet there remains no algorithm for the spark of recognition which fires when we see a friend’s face.

At Mugr we wonder about not only how to better answer those fundamentally human questions, but how to make even our small strides useful to the people around us. We’ve experimented with the notion of direct image search: enter a photo, find the person. We were surprised at the amount the public bristled at the potential invasion of privacy, but learned there might be more useful things we could attempt. Our current focus centers on using face recognition to let people automatically index and search their online photo galleries. It seems to resonate more positively with the people around us, but we’ve found it’s certainly no less challenging. Recently, we’ve begun to look at humanitarian applications for our technology. It’s been incredible to learn that we might be able to make a larger difference in the world.

I can’t say if we’ll make tremendous progress on these hard problems; maybe that’s for another group to accomplish. I can’t say if we’ll manage to make something that improves the lives of the people around us; perhaps that’s the realm of other companies. What I can say is that we’re learning– about far more than we ever thought we would– and enjoying the challenge.

This entry was posted on Wednesday, February 6th, 2008 at 2:10 pm and is filed under Alts, CEO Views. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.