A novel YOLO26-MoE optimized by an LLM agent for insulator fault detection considering UAV images
Researchers have developed a new object detection model, YOLO26-MoE, to improve the automated inspection of electrical power line insulators using UAVs. This model integrates a Mixture-of-Experts (MoE) module to better refine features for detecting subtle and varied fault patterns. An LLM agent was utilized to coordinate the optimization and training process, resulting in state-of-the-art performance with a 0.9900 [email protected]. AI
IMPACT Introduces an LLM-optimized model for improved infrastructure inspection, potentially enhancing grid reliability.