JUST no chance to travel incognito even for bears and maybe, all animals in the very near future. Because Facial Recognition is getting better and now helping researchers keep track of bears.

Ed Miller and Mary Nguyen are Silicon Valley software developers by day, but moonlight at solving unusually fuzzy problems.

A few years ago they became mesmerized like many of the rest of us, by an Alaskan webcam broadcasting brown bears from Katmai National Park. They also happened to be seeking a project to hone their machine learning expertise.

''We thought machine learning is really great at identifying people; what could it do for bears?'' Mr.Miller said. Could artificial intelligence used for face recognition be harnessed to distinguish one bear from another.

At Knight Inlet in British Columbia, Canada, Melanie Clapham was pondering the same question.

Dr. Clapham, a postdoctoral researcher at the University of Victoria working with Chris Darimont of the. Raincoast Conservation Foundation, was eager to explore  facial recognition technology as an aid in her grizzly bear studies .But her expertise was bear biology, not A.I.

Fortunately, the four found a match on Wildlabs.Net, an online broker of collaborations between technologists and conservationists. Combining their skill sets, Mr. Miller and Ms. Nguyen volunteered spare time over several years for this passion project, reporting the results of their experiment last week in the journal Ecology and Evolution.

The project they produced, BearID, could help conservationists monitor the health of bear populations in various parts of the world and perhaps aid work with other animals, too.

They got started by looking for other animals that had gotten the deep learning treatment.

''In typical engineering fashion, we're looking for a shortcut,'' Mr. Miller said.

They discovered ''dog hipsterizer,'' a program that found the faces, eyes and noses of dogs in photos and placed rimmed glasses and mustaches on them. ''That was where we started,'' Ms. Nguyen said.

Although trained on dogs, dog hipsterizer worked reasonably well on the similarly shaped faces to of bears giving them a programming head start. Nevertheless, Ms. Nguyen said the work's initial stages were tedious.

Creating a training data set for the deep learning program involved examining over.

4,000 photos with bears in them and then manually highlighting each bear's eyes, nose and ears by drawing boxes around them so the program could learn to find these features.

The system also had top overcome a challenge of brown bears physical appearance.

To monitor populations, ''we have to be able to recognize individuals,'' said Dr. Clapham. But bears don't have any feature capable to a fingerprint, such as a zebra's stripes or giraffe's spots.

From 4,675 fully labelled bear faces on photographs taken from research and at bear viewing sites at Brooks River, Alaska and Knight Inlet, they randomly split images into training and testing data sets.

Once trained using 3,740 photographs of bear faces, deep learning went to work ''unsupervised,'' Dr. Clapham said, to see how well it could spot differences between known bears from 935 photographs.

The honor and serving of the latest global operational research on A.I., Facial Recognition and Future continues to Part 2. The World Students Society thanks author Lesley Evans Ogden.


Post a Comment

Grace A Comment!