An open-source project dedicated to tracking and segmenting any objects in videos, either automatically or interactively. The primary algorithms utilized include the Segment Anything Model (SAM) for key-frame segmentation and Associating Objects with Transformers (AOT) for efficient tracking and propagation purposes.
The background / unlabeled part of a mask is index 0.
If you have two objects being tracked, they would be index 1 & 2 and
SegTracker.curr_idx = 2
Then when SegTracker.find_new_objs() finds one additional object, it should set the index in the resulting combined mask to SegTracker.curr_idx + 1 = 3
Currently the increment happens after assignment, so the third object incorrectly gets assigned SegTracker.curr_idx = 2
Please let me know if this is correct:
The background / unlabeled part of a mask is index 0.
If you have two objects being tracked, they would be index 1 & 2 and
SegTracker.curr_idx
= 2Then when SegTracker.find_new_objs() finds one additional object, it should set the index in the resulting combined mask to
SegTracker.curr_idx
+ 1 = 3Currently the increment happens after assignment, so the third object incorrectly gets assigned
SegTracker.curr_idx
= 2This might be the cause of https://github.com/z-x-yang/Segment-and-Track-Anything/issues/105