Individual identification of brown bears using pose-aware metric learning

Identifying individuals within a species is a cornerstone of ecological and biological research, holding the key to understanding behavior, population dynamics, and conservation needs. However, conventional identification methods such as genetic sampling and physical tags come with significant trade-offs, including invasiveness, high costs, and limited scalability. Despite remarkable advances in computer vision that have transformed the identification of patterned species, reliably distinguishing individuals in unmarked species remains an open challenge. Here, we curate a novel dataset of visually identified individual Alaskan coastal brown bears (Ursus arctos). This 72,940-image dataset contains high-resolution images of 109 known individuals from multiple seasons and in varied conditions (e.g., fur shedding, substantial weight gain). Identification of these bears is notoriously difficult, yet our new pose-aware metric-learning-based AI model is able to leverage biometric information that enables individual re-identification (ReID) with promising accuracy for future research and conservation efforts. Our findings indicate that it is possible to reidentify individual brown bears across years and to detect unknown individuals in a "real-world" open dataset. PoseSwin presents a promising approach to the challenge of non-invasive ReID of other unmarked species, as well, and points toward an expanded range of possible questions in wildlife and ecological research.