Abstract: Edge intelligent visual monitoring of substations performs intelligent detection on collected images at edge intelligent perception terminals, representing a novel substation intelligent ...
Abstract: Large Vision-Language Models (LVLMs) suffer from severe object hallucinations, leading them to frequently generate outputs that do not correspond to the image content, significantly reducing ...