In LaSOT, we conduct one-pass evaluation (OPE) to assess the performance of each tracker. In detail, we utilize three types of metrics, Precision, Normalized Precision and Success, to measure different tracking algorithms. The definitions of the three metrics can be seen in the paper.
We define two protocols for evaluating trackers on LaSOT as follows
We assess 35 popular tracking algorithms on LaSOT under Protocol I and II. These trackers include deep learning based ones, correlation filter based ones with hand-crafted or deep features, sparse representation based ones and other representatives. Table 1 shows these trackers.
Note: Each tracker is used as it is from authors' implementation, without any modification.
The following plots demonstrate the evaluation results of 35 trackers under two protocols using three metrics.