The moods of animals are known to affect the quality of output of their products. To understand animalistic behaviours and moods better, an animal scientist from Netherland's Wageningen University & Research has built an artificial intelligence (AI) based application to study and understand the emotional state of farm animals based on their photographs taken via a smartphone. Suresh Neethirajan, the animal scientist, created the AI application, WUR Wolf (Wageningen University & Research: Wolf Mascot), to respond and measure emotions to provide a natural user interface to enable the digitalization of animal welfare platforms. The application is a real-time facial expression recognition platform that can automatically code the emotions of farm animals.

"Using Python-based algorithms, we detect and track the facial features of cows and pigs, analyze the appearance, ear postures, and eye white regions, and correlate with the mental/emotional states of the farm animals," says Neethirajan in the blog on biorxiv.org. 

There is scientific evidence that states that animals under less stressful environments are more productive, such as finding lesser stress-related hormones in meat, etc. Recent scientific research also suggests that adapting to their natural environments, where they can be free and playful, can yield better results. Happy cows or goats, for example, are likely to produce more milk than those that are bored.

Neethirajan has trained his application on datasets of animal's facial expressions from over six farms. So far, the application is able to detect 13 facial actions and 9 emotional states of animals with an average accuracy of 85%.

"The software detects 13 facial actions and 9 emotional states, including whether the animal is aggressive, calm, or neutral...Detecting expressions of farm animals simultaneously in real-time makes many new interfaces for automated decision-making tools possible for livestock farmers. Emotions sensing offers a vast amount of potential for improving animal welfare and animal-human interactions," concludes Neethirajan.