Precise Annotated data
YangBain opened this issue · comments
@hepek @marcovzla @Nikhil-Kasukurthi @cvondrick Recently, I use vatic to annotate video data. But when I use F and D for forward and backward, the results are not very good. So, in this case, who knows how to get the precisely annotated data. And if I use V and C directly, I can not save the results. So could you give me some suggestions for how to get the precisely annotated data. Thank you very much.
@YangBain I have exact the same problem. Did you figure out how to solve it ?
Edit:
I solved the problem. The solution is to change track.js file from vatic public directory to track.js file from @onauparc pull request #20. What is worth to notice is to refresh browser (shift+f5), after changing track.js files.
@gosowski I am sorry I do not solve this problem. Finally, I use another image annotation tool to do annotation work. Vatic always output many wrong bounding boxes for my case.
another image annotation tool ？What tool?