
Combining Vision-Language Models with graph representations, HGCLIP enhances hierarchical image classification by integrating hierarchical relationships and multi-modal prompts.


coremsg

hgclip-vision-language-models-with-graph-representations-for-hierarchical-understanding


HGCLIP: Vision-Language Models with Graph Representations for Hierarchical Understanding


title_rewrite


組織されたカテゴリを多階層の分類階層に探索する新しい方法を提案します。


hgclip-exploring-vision-language-models-with-graph-representations-for-hierarchical-understanding


HGCLIP: Exploring Vision-Language Models with Graph Representations for Hierarchical Understanding
