|
Eurek-baa: Scientists find sheep can
recognize human faces
Send a link to a friend
 [November 09, 2017]
LONDON (Reuters) - Sheep have been
trained to recognize the faces of celebrities, including former U.S.
President Barack Obama, by University of Cambridge scientists who hope
it may help with understanding neurodegenerative diseases.
In a specially equipped pen, sheep were shown pictures of people on two
computer screens, on one side would be an unknown person and on the
other would be one of four celebrities.
The animal would receive a reward of food for choosing the photograph of
the celebrity by breaking an infrared beam near the screen displaying
it. If they chose the wrong photograph, a buzzer would sound and they
would receive no reward.
The sheep eventually managed to identify the familiar face eight times
out of every 10.
The group of celebrities the sheep were trained to recognize included
actors Emma Watson and Jake Gyllenhaal, BBC newsreader Fiona Bruce and
Obama.
"We've shown that sheep have advanced face-recognition abilities,
comparable with those of humans and monkeys," Professor Jenny Morton,
who led the study, said in a statement.
In addition to being shown images of the celebrities facing forward,
scientists also tested the animals' ability to recognize the faces in
photographs taken from other angles.
[to top of second column] |
A picture taken from undated video shows a sheep approaching a photo
of Barack Obama during a Cambridge University experiment, in
Cambridge, Britain. Jenny Morton/University of Cambridge handout via
REUTERS
The animals' success rate fell by around 15 percent when presented
with the faces at a new angle, an amount researchers said was
comparable to that seen when humans perform the task.
Scientists aim to use the sheep as models to understand disorders of
the brain, such as Huntington's disease, that develop over a long
time and affect cognitive abilities.
(Writing by Mark Hanrahan in London Editing by Jeremy Gaunt)
[© 2017 Thomson Reuters. All rights
reserved.]
Copyright 2017 Reuters. All rights reserved. This material may not be published,
broadcast, rewritten or redistributed.
 |
