F-LSeSim icon indicating copy to clipboard operation
F-LSeSim copied to clipboard

Published 20 hours ago • verified publisher icon lyndonzheng

Selecting query-key patches

Open YangGangZhiQi opened this issue 2 years ago • 2 comments

@lyndonzheng Hi, thanks for your great work. After reading your paper and source code, I have a question about selecting query-key patches. In your source codes, the selecting codes are as below. selecting_patch In line 246, left, top = pos_x - int(pw/2), pos_y - int(ph/2). What's the reason to subtracts int(pw/2) from pos_x ? Watching codes from line 244 to line 259, It seems like a strategy to select query-key patches. Is there some reason to use this strategy or we could use other strategy to get same results ?

YangGangZhiQi avatar Jul 06 '22 02:07 YangGangZhiQi

Hi @YangGangZhiQi, thanks for your interesting. This strategy is applied to ensure the valid sample patches. Without this processing, the patch may be out of the original feature scale.

lyndonzheng avatar Aug 09 '22 03:08 lyndonzheng

Get it. Thanks for your quick answer. These days a question often troubles me when working in Image-to-Image transfer field. How to explain the correctness of the generated results? Is there some other way except just observation to declare there is no error in the generated images? I would appreciate it if someone could help me.

YangGangZhiQi avatar Aug 26 '22 07:08 YangGangZhiQi