Identifying human facial expressions that convey emotions is easy. Faces that show sorrow, pain, surprise, anger, and joy can be easily identified. The same cannot be said when it comes to animals. Charles Darwin might be correct in his proposal that the universal way emotions are communicated in animals and humans is through facial expressions, but it will be a pain in the butt to identify animal facial expressions, as analysis of these things will be subject to bias in human scoring. Thankfully, we have machine learning that can help us.
By applying machine learning to the facial expressions of mice, researchers at the Max Planck Institute of Neurobiology, Germany, were able to identify distinct emotional states of pleasure, disgust, nausea, pain and fear, and even their relative strength.
Importantly, the study showed these weren’t merely a reaction to a sensory stimulus, but reflected the inner, underlying emotion.
More details about this study over at Cosmos Magazine.
(Image Credit: Julia Kuhl/ Cosmos)