What are the details of using the visual grouding tool? about hvpnet HOT 2 CLOSED

Hi,

Thanks for your attention! Your question may be answered as follows: :
(1) If there is more than one noun, each noun will be queried separately.
(2) During the extraction process, we will consider the head entity nouns and tail entity nouns in the text. Specifically, we will utilize these entity nouns to extract image regions in visual grounding tools, resulting in distinct image regions for the data with the same text and the same image but different head entity or tail entity.

from hvpnet.

Thank you!

from hvpnet.