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 automatically identifies common marmosets in real time with over 82.9% precision and 91.5% recall, matching human-level performance.

• Why it matters: Accurate, real-time animal identification is crucial for studying behavior, neural activity, and health, but current methods are invasive, time-consuming, and variable across experimenters.
• What they did: The authors developed an automatic pipeline using supervised deep learning with YOLOv8 to localize, detect faces, and classify identities of marmosets, enhanced by color-coded collar bead recognition, tested across adult and juvenile groups.
• The result: This system enables rapid, non-invasive, and generalizable identification across species, ages, and hardware, facilitating more efficient behavioral and physiological research in non-human primates.