Zihe Huang

Differentiated Vision: Unveiling Entity-Specific Visual Modality Requirements for Multimodal Knowledge Graph

Multimodal Knowledge Graphs (MMKGs) enhance knowledge representations by integrating structural and multimodal information of entities. Recently, MMKGs have proven effective in tasks such as information retrieval, knowledge discovery, and question answering. Current methods typically utilize pre-tra