Cow vocalizations carry important cues about a herd’s welfare and needs, making them valuable for health monitoring in precision livestock farming. We present MooLogue, a cross-platform mobile application that offers farmers and researchers an interactive interface to explore annotated cow calls and their meanings. The app combines a Flutter front-end with an Amazon Web Services (AWS)-based, Python/PyTorch backend that is used to generate and manage labelled audio examples of cow vocalizations. By providing a curated call library of cow call recordings linked to behavioural classes (call types), grouped into nine categories, model confidence scores, and interpretive text, MooLogue bridges bovine bioacoustic research with on-farm learning and welfare-oriented decision-making.
Kate et al. (Thu,) studied this question.