ComfyUI Node

SemanticImageFetch

Select the closest images to an input prompt

By yasser-baalla·Created about a year ago·Updated about a year ago· 0
SemanticImageFetch
  • image
  • clip
  • clip_vision
  • IMAGE
prompt
number_of_candidates1
CategoryIMAGE

Inputs (5)

NameTypeDefaultDescription
imageIMAGEThe list of images to fetch the semantic map from.
clipCLIPThe CLIP model used for encoding the text.
promptSTRING
clip_visionCLIP_VISIONThe CLIPVision model used for encoding the images.
number_of_candidatesINT11–10Number of closest images to retrieve.

Outputs (1)

NameTypeDescription
IMAGEIMAGE