This study compares the effectiveness of Artificial Intelligence (AI) against human experts in assessing acute pain in sheep, utilizing video recordings and photographs from 48 sheep undergoing surgery. The research specifically evaluates AI’s performance using facial expression analysis (SFPES) and behavioral scoring (USAPS), which is considered the “golden standard.” Findings indicate that the AI significantly outperformed humans in facial pain recognition and equaled human performance in behavioral pain assessment, despite only having access to facial images. The authors suggest these results have substantial implications for clinical veterinary practice, proposing that AI could transform animal pain assessment by offering a more objective and less biased method than traditional human scoring.
Supervised Neural Style Transfer as an Augmentation Technique for Facial Landmark Detection
A major challenge in training accurate facial landmark detection models
