Object Detection for Efficient Lighting: Comparing Models

Light energy saving systems are important in different spheres of life and help in increasing energy efficiency, reducing costs and sustainability promotion. Despite their wide use, classical versions have various challenges such as poor granulation, problems with anomaly detection, lack of precision in data reporting, high costs of sensors and difficulties related to integration. This research paper presents an innovative light energy saving system that performs a comprehensive evaluation of YOLO; faster RCNN; and SSD models to determine the most effective solution for enhancing efficiency. The purpose of this study is to provide end-users with real-time understanding about their energy usage habits. Based on a dedicated dataset and rigorous testing, the project seeks to identify the best model type. Selected models then go through a test using CCTV cameras footage to establish which lights devices are unnecessarily left on. The IoT automation also operates alongside it hence deactivating discovered appliances. It has adopted a holistic approach that not only drives energy conservation but also promotes cost saving and sustainable light energy consumption within its own context. This research report is to develop light energy saving systems through introduction of a model-agnostic approach. The study aims at identifying the most efficient deep learning model, through meticulous comparative analysis, that will address the limitations of conventional systems and promote more sustainable and cost-effective solution for real-time energy optimization.