A real-time, multi-animal model for automatic face detection and identification of freely moving common marmosets based on YOLOv8 algorithms.

A YOLOv8-based system achieves over 82.9% precision and 91.5% recall in real-time identification of freely moving common marmosets, matching human accuracy.

• Why it matters: Accurate, non-invasive, and real-time animal identification is crucial for studying behavior, neural activity, and health, but current methods are time-consuming and disruptive.
• What they did: The authors developed an automatic pipeline using deep learning and collar bead recognition to localize, detect faces, and identify individual marmosets in real-time from video.
• The result: This system enables rapid, automatic, and generalizable identification across ages and species, facilitating more efficient behavioral and physiological research.