How do you get the reasonable subset from the imageset ?

SoonminHwang / rgbt-ped-detection

KAIST Multispectral Pedestrian Detection Benchmark [CVPR '15]

http://multispectral.kaist.ac.kr

Other

304 stars 61 forks source link

How do you get the reasonable subset from the imageset ? #8

Closed anushl9o5 closed 5 years ago

anushl9o5 commented 5 years ago

I have downloaded the dataset and the image sets as well. I also get that the subset is obtained by sampling every 20 images from the test set. But what I don't get is on what basis do you ignore a bounding box?