HumanParsing-Dataset icon indicating copy to clipboard operation
HumanParsing-Dataset copied to clipboard

Published 20 hours ago • verified publisher icon lemondan

No detail information for the data

Open sizhangyu opened this issue 8 years ago • 1 comments

Could you please provide some detail information about how to use this dataset? For example, each "image" is a 1-d vector with different size. How should we reconstruct the image? What do "index" and "id" mean? All "url" is empty? Thank you.

sizhangyu avatar Aug 05 '16 06:08 sizhangyu

you can use the decoding code in http://vision.is.tohoku.ac.jp/~kyamagu/research/paperdoll/

lemondan avatar Aug 05 '16 13:08 lemondan