Application of YOLOv11 deep learning model for classification and counting ice-rafted debris (IRD) in core sediments in the Arctic Ocean

Key Points

• Classification of ice-rafted debris showed significant accuracy, enhancing understanding of Arctic sediment dynamics.
• The YOLOv11 deep learning model achieved over 90% accuracy in identifying ice-rafted debris types.
• Assessment utilizing advanced deep learning methods demonstrates the potential for automated counting of debris in core sediments.
• Findings imply a novel approach for monitoring environmental changes in the Arctic, highlighting its ecosystem sensitivity.