Researchers Design Facial Recognition System for Lemurs

91 views Leave a comment

University of Arizona anthropologist Stacey Tecot has spent 17 years study red-bellied lemurs in Madagascar’s Ranomafana National Park. One of a biggest ongoing hurdles in conducting her research: gripping lane of who’s who in a lemur community.

In hunt of a approach to improved lane and commend particular lemurs in a furious – in a slightest invasive approach probable – Tecot and associate anthropologist Rachel Jacobs, of George Washington University, enlisted a assistance of their co-worker Andrea Baden and mechanism scientists to emanate a facial approval complement means of identifying lemurs by their particular earthy characteristics.

LemurFaceID recognizes lemurs’ skin and fur patterns with scarcely 99 percent accuracy. Image credit: Stacey Tecot

The outcome was LemurFaceID, a computer-assisted approval complement that has a intensity to redefine how researchers lane opposite class in a wild.

The technology, described in a stream emanate of a biography BMC Zoology, will support researchers in critical evolutionary studies and assist in charge efforts for lemurs, that were named a world’s many involved organisation of mammals in 2012.

Researchers have typically relied on one of dual methods to lane lemurs. One is to physically constraint and tab them – something Tecot and Jacobs have avoided, given it can be stressful for a animals. The other is to simply observe a animals visually, that takes a lot of legwork and requires rarely learned and gifted researchers for a many accurate results.

With LemurFaceID, researchers will be means to build a database of photos of Madagascar’s lemurs, that they can afterwards use to brand particular animals and to assistance endorse their margin observations. The record could also make it easier for researchers to combine and share information and control some-more integrated research.

“Studying people and populations over prolonged durations of time provides essential information on how prolonged people live in a wild, how frequently they imitate as good as rates of tot and youthful mankind and eventually race expansion and decline,” pronounced Tecot, an partner highbrow in a UA School of Anthropology and comparison author of a research. “Information like that can surprise charge strategies for lemurs, a rarely involved organisation of mammal.”

LemurFaceID works on beliefs identical to tellurian facial approval systems, like a one we confront when tagging friends on Facebook.

In a box of lemurs, it recognizes facilities like hair and skin patterns, and it works with considerable 98.7 percent correctness when given dual face images of a animal.

Researchers wish a module can offer as a indication for tracking other class and, in some cases, potentially reinstate physically tagging animals.

“We consider this process could be practical to studies of class that have identical movement in hair and skin patterns, such as red pandas and some bears, among others,” pronounced Jacobs, a biological anthropologist during George Washington University’s Center for a Advanced Study of Human Paleobiology and co-lead author of a research.

Tecot, Jacobs and Baden worked with mechanism scientists during Michigan State University and Hunter College in New York to rise and exam LemurFaceID regulating a dataset of 462 images of 80 particular red-bellied lemurs and a database of 190 images of other lemurs. There are over 100 opposite lemur class found in Madagascar.

Tecot pronounced a facial approval complement could assistance researchers do a consummate race census to reassess a standing of red-bellied lemurs in Madagascar. The record also might assistance her and Baden with their stream research, that focuses on a expansion of common tot caring in red-bellied lemurs and because some fathers and siblings share tot caring responsibilities, while others don’t.

Tecot hopes a facial approval complement can one day be extended to an educational smartphone focus that would concede locals and tourists in Madagascar to compare photos they take of furious lemurs with images in a database to learn some-more about a particular animal.

“We see lots of opposite intensity applications for this,” Tecot said. “This is only a initial step for us in holding this in many directions.”

Source: University of Arizona

Comment this news or article