Chick call detection for animal–computer interaction system promoting growth of chicks

Vocal communication between hens and chicks promotes the growth of chicks by decreasing their fear response and increasing their feeding behavior. However, poultry farms raise chicks without hens for production efficiency reasons. Chicks raised in such an environment react strongly to the movements or sounds of breeding staff, crowding and hurting each other, leading themselves to a negative impact on growth. To solve this problem, conventional approaches employ chick–computer interaction system simulating the hen's vocal communication with her chicks which is known to reduce the fear responsiveness of chicks to novel objects. This suggests that chick–computer interaction system potentially alleviates overreactions of chicks to external stimuli and promotes their growth, which is important for improving animal welfare in livestock. In order to implement such a system in poultry farms, it is necessary to accurately detect the chicks' call under environmental including the sound of footsteps or ventilation fan. In this study, we propose a method for detecting chicks' call that utilize the frequency characteristics of chicks' call and suppresses the false detection of noise. Furthermore, we have developed the real-time system responding to chicks' call that indicates negative emotion.