Results
The processed videos using the proposed system are presented below. For each sequence, the inital object position has been selected manually drawing the initial bounding box.
Long Term sequences:
Challenges: