Automated Dress Code Compliance Monitoring using YOLO based Models

Dress code detection in diverse settings including educational institutions, industrial environments and public spaces has become increasingly relevant to automated detection of dress code. In this review, which presents a comparative analysis of YOLO-based object detection models advancements in dress code compliance. The models that start with YOLOv5, transition to YOLOv11 and YOLO-NAS have significantly improved in accuracy, compute efficiency, and real time performance. In general settings, YOLOv10 is a great tradeoff between speed and precision. At the same time, YOLO-NAS proves to be the most accurate and the most efficient model in the industrial environments, able to detect safety attire under more challenging conditions. These models redefine the potential of automated monitoring systems by the integration of innovations such as Neural Architecture Search (NAS), spatial attention mechanisms, and advanced feature extraction techniques. Additionally, this review features key metrics including precision, recall, Mean Average Precision (mAP) and latency, making it easy to learn from them to select the most appropriate model for downstream dress code detection scenarios. Future research aims to strike a balance the tradeoff between resource efficiency and adaptability by developing lightweight models for edge devices to be run on, and increasing the adaptability to accommodate more complex and dynamic environments.