Person Centric Visual Grounding

Person-centric Visual Grounding refers to the task of associating the person mentioned in a textual description with the actual person appearing in an image. This task aims to achieve precise localization and recognition of specific individuals by integrating visual and textual information, thereby enhancing the accuracy and efficiency of multimodal content understanding. This technology has significant application value in fields such as human-computer interaction, intelligent surveillance, and multimedia information retrieval.