Closed GOBish closed 4 years ago
I am trying to use my object identifier (a tensorflow model) to create the detections for SORT but I am having some trouble formatting it correctly.
I am following the advice in https://github.com/kedpter/track_face_example/blob/master/track_face_example.py from https://github.com/abewley/sort/issues/72
with something similar:
dets_list = [l, t, r, b, 1] #I get this from the output of my detector dets = np.array(dets_list) trackers = mot_tracker.update(dets)
however when I do this I get an error from SORT
IndexError: too many indices for array trk=KalmanBoxTracker(dets[i,:])
Note this is when there is only one detection in the frame - i think that may be causing the issue.
Ahh, just had to add another bracket.
dets_list = [[l, t, t, b, 1]]
I am trying to use my object identifier (a tensorflow model) to create the detections for SORT but I am having some trouble formatting it correctly.
I am following the advice in https://github.com/kedpter/track_face_example/blob/master/track_face_example.py from https://github.com/abewley/sort/issues/72
with something similar:
however when I do this I get an error from SORT
Note this is when there is only one detection in the frame - i think that may be causing the issue.