Public Release: 

Here's looking at you

Dogs can discriminate emotions in human faces

University of Veterinary Medicine -- Vienna

Dogs can discriminate human faces on pictures. This was demonstrated by Ludwig Huber and colleagues at the Messerli Research Institute in 2013.

But are dogs also able to discriminate human emotional expressions? To date, no study could answer this question convincingly.

Dogs discriminate human emotions on a touchscreen

Corsin Müller and Ludwig Huber from the Messerli Research Institute investigated this skill together with colleagues at the Clever Dog Lab at the Vetmeduni Vienna. They presented photos of happy and angry women's faces side by side on a touchscreen to 20 dogs.

During the training phase, dogs from one group were trained to touch images of happy faces. The other group was rewarded for choosing angry faces.

To exclude the possibility that the animals were making their decisions based on conspicuous differences between the two pictures, such as teeth or frown lines, the researchers split the images horizontally so that during the training phase the dogs saw either only the eye region or only the mouth region. Most of the dogs learned to differentiate between the happy and angry face halves. They subsequently also managed to identify the mood in novel faces as well as in face halves that they had not seen during the training phase.

Approaching happy faces is easier

Dogs trained to choose the happy faces mastered the task significantly faster than those who had to choose the angry faces. "It seems that dogs dislike approaching angry faces," study director Ludwig Huber explains.

"We believe that dogs draw on their memory during this exercise. They recognize a facial expression which they have already stored," first author Corsin Müller explains. "We suspect that dogs that have no experience with people would perform worse or could not solve the task at all."

Dogs' visual abilities are underestimated

Dogs have a much better sense of smell and hearing than humans, but the spatial resolution of their vision is about seven times lower. "It had been unknown that dogs could recognize human emotions in this way. To better understand the development of these skills, we want to perform similar tests also with wolves at the Wolf Science Center," Huber says.

As part of the WWTF project "Like me", the team led by Ludwig Huber has spent the past three years investigating whether dogs are able to understand the emotions of conspecifics or of people. Project partners at the University of Vienna and the Medical University Vienna are performing similar research on the empathic abilities of people.

###

Service:

The article "Dogs Can Discriminate Emotional Expressions of Human Faces", by Corsin A. Müller, Kira Schmitt, Anjuli L. A. Barber and Ludwig Huber will be published on the 12th of February 2015 at 12:00 EST (18:00 MEZ) in the journal Current Biology.

About the University of Veterinary Medicine, Vienna

The University of Veterinary Medicine, Vienna in Austria is one of the leading academic and research institutions in the field of Veterinary Sciences in Europe. About 1,300 employees and 2,300 students work on the campus in the north of Vienna which also houses five university clinics and various research sites. Outside of Vienna the university operates Teaching and Research Farms. http://www.vetmeduni.ac.at

Scientific Contact:

Prof. Ludwig Huber
Messerli Research Institute - Unit of Comparative Cognition
University of Veterinary Medicine Vienna (Vetmeduni Vienna)
ludwig.huber@vetmeduni.ac.at

or

Corsin Mueller
Messerli Research Institute - Unit of Comparative Cognition
University of Veterinary Medicine Vienna (Vetmeduni Vienna)
T +43 1 25077-2688
corsin.mueller@vetmeduni.ac.at

Released by:

Susanna Kautschitsch
Science Communication / Public Relations
University of Veterinary Medicine Vienna (Vetmeduni Vienna)
T +43 1 25077-1153
susanna.kautschitsch@vetmeduni.ac.at

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.