LUB-YOLO: A lightweight method for laboratory unsafe behavior detection based on improved YOLOv8s

Key Points

• Detection of unsafe behavior achieves 92% precision and 88% recall, significantly enhancing lab safety.
• Utilization of an improved YOLOv8s model demonstrates rapid identification of unsafe actions in real-time.
• Machine learning algorithms are employed to analyze video footage, ensuring reliable safety monitoring.
• This approach highlights the need for continuous assessment and real-time monitoring to prevent accidents.