Skip to contents

Model-free: give it the embeddings and it applies each feature's rule. centroid_patch_max features need patches; clip_text_pair features need image and textEmb (rows named by prompt). For external features supply the scores yourself.

Usage

featureScores(key, image = NULL, patches = NULL, textEmb = NULL)

Arguments

key

A dimoose key with a features table.

image

objects x d matrix of image embeddings (unit rows).

patches

objects x patches x d array of patch embeddings.

textEmb

prompts x d matrix of text embeddings whose row names are the prompt texts, e.g. from embedTexts().

Value

An objects x features matrix with feature ids as column names.

See also

scoreImages() for images-to-scores in one call.