VISO icon indicating copy to clipboard operation
VISO copied to clipboard

Published 20 hours ago • verified publisher icon QingyongHu

questions about SOT dataset

Open zhaoxingle opened this issue 2 years ago • 0 comments

Hello, author. Recently I visualized the ground-truth images in the single object task test set 24 to 27, and found that some annotations are not very accurate, such as the first frame does not contain objects S0 %~ _LH RS_Z AX_R3QR , or there are some frames in the middle, the annotations directly become all 0, 5N OQYT15B)IOJY2 LK6E~Y and this There is nothing in the first frame frame, and there is no object in the subsequent frame, $@2W) J}JWJ7)P}FM)PGQ J and there is nothing in the whole sequence. There are also ground-truth, proving that my visualization code is fine. {CX}1O_GKA9R5ZWG)3 54 _AT6C2S$DB)4V7`H{1R5LOI I would like to take the liberty to ask the author if something went wrong?

zhaoxingle avatar Oct 21 '22 02:10 zhaoxingle