Time Is Running Out For Lemurs. Is Facial Recognition Tech A Last Resort?
Lemurs — small primates endemic to Madagascar — are among the most endangered mammals on Earth due to hunting and the destruction of their tropical forest habitat by illegal loggers.
Researchers say facial recognition technology could not only help inform strategies for conserving lemur populations, but also allow us to explore demographic and evolutionary processes that shape lemur populations.
A new computer-assisted recognition system called LemurFaceID uses facial characteristics of lemurs from photographs taken in the wild to make positive identifications of individual animals.
Lines of research facilitated by LemurFaceID are manifold: individual lemurs can be tracked over time, records of how many individuals there are in any given population can be compiled, and the social systems of those populations can be more closely examined.
Lemurs were named “the world’s most endangered mammal” in 2012, and their fortunes have not improved much since then. Four lemur species were included in a list of the 25 most endangered primates assembled by the International Union for the Conservation of Nature (IUCN) in 2015, for instance.
“Our results indicate that fully 94 percent of the lemurs of Madagascar are threatened, by far the highest percentage for any larger group of mammals, with 66 percent of these falling into the Critically Endangered and Endangered categories,” said Russell A. Mittermeier and Anthony B. Rylands, who serve on the IUCN Species Survival Commission Primate Specialist Group.
Any new technology that can aid in lemur conservation efforts would be welcome. In a recent paper published in the journal BMC Zoology, a team of researchers said that facial recognition software could be such a potent tool that they might not only help inform strategies for conserving lemur populations but could also allow us to explore the demographic and evolutionary processes that shape lemur populations more broadly.
A team of researchers has developed a new computer-assisted recognition system that can identify individual lemurs in the wild by their facial characteristics and ultimately help to build a database for long-term research on lemur species.
“The ability to consistently study individuals over long periods of time, as well as integrate data across different studies, are some of the challenges we face when studying wild animal populations,” said Rachel Jacobs, a biological anthropologist at the George Washington University’s Center for the Advanced Study of Human Paleobiology, in a statement. “Senior author, Stacey Tecot (University of Arizona), and I weren’t particularly satisfied with the common approaches used in lemur research, so we aimed to do something different with red-bellied lemurs, and we sought the expertise of our computer science collaborators.”
The new computer-assisted recognition system Jacobs and team developed is called LemurFaceID. The researchers say it can use facial characteristics of lemurs from photographs taken in the wild to make positive identifications of individual animals.
Jacobs and her co-authors modified human facial recognition technology in order to identify individual lemurs based on variations in their facial patterns. “LemurFaceID was able to identify individual lemurs based on photographs of wild individuals with a relatively high degree of accuracy,” they write in the paper. “This technology would remove many limitations of traditional methods for individual identification. Once optimized, our system can facilitate long-term research of known individuals by providing a rapid, cost-effective, and accurate method for individual identification.”
The lines of research facilitated by LemurFaceID are manifold, according to the paper: individual lemurs can be tracked over time, records of how many individuals there are in any given population can be compiled, and the social systems of those populations can be more closely examined. It could also even help track lemurs when they’re taken from their natural habitat by wildlife traffickers.
Evolutionary studies require long-term data on population dynamics, which typically involve capturing and tagging individual animals and then releasing them back into the wild in order to track metrics on mortality rates, population growth, and reproduction. Because of how costly and time-consuming this can be, large-scale, population-level studies are undertaken for very few species.
“Long-term data needed to address evolutionary questions are lacking for many species,” the researchers write in the paper. “This is, at least in part, due to difficulties collecting consistent data on known individuals over long periods of time.”
Taking photos of lemurs and uploading them to LemurFaceID can accomplish this task in a much less-invasive manner and in a much shorter time than traditional methods such as tagging, which can involve the use of nets, traps, and sometimes even blow guns or air rifles, the researchers say.
“Capturing has several advantages,” the researchers note in the paper, “such as enabling data to be collected that would otherwise be impossible (e.g., blood samples, ectoparasites), but it can also be expensive, often making it unfeasible for studies with large sample sizes and/or those conducted over large spatial and temporal scales.”
Jacobs and her co-authors say that software like LemurFaceID could be employed to track other species, potentially even replacing the work of physically tagging individual animals altogether for some populations.
“We think this method could be applied to studies of species that have similar variation in hair and skin patterns, such as red pandas and some bears, among others,” Jacobs said.
Read more at Mongabay.