Fig. 4
From: Image region semantic enhancement and symmetric semantic completion for text-to-image person search

Results of superpixel visual patch division.
From: Image region semantic enhancement and symmetric semantic completion for text-to-image person search

Results of superpixel visual patch division.