Inconsistency? in train_online.py

[nijkah]

in Line 79 of train_online.py,
It generates DAVIS 2017 loader.
Although it requires all images and annotations of each sequence,
DAVIS2017-test doesn't provide all annotations but first frame.

How should I change the code? or path?

[colorblank]

I come across the same problem.Did you solve it?

[aaaaannnnn11111]

I come across the same problem.Did you solve it?

Did you both solve the problem?

[nijkah]

I just wrote new code about dataset and training.
You can check in here.
https://github.com/nijkah/MaskTrack_Box