Questions about building Cops-Ref dataset
Buki2 opened this issue · comments
Hi! Thanks for sharing the data. I really enjoy this excellent work :)
I'm a little confused about the expression engine. How does the engine choose the logic form for a given region? I mean different scenes suit different logic forms. It's not likely to be randomly chosen, is it?
I also want to know how to get the distracting images especially the most difficult type Cat&cat. Is it manually get?
Thank you!