To participate in the MCT challenge, the participant is required to conduct at least one experiment. But if you want to put the tracking results in the ranking, all the three experiments must be conducted. In each experiment, the tracker should be separately tested on the four subsets in the NLPR_MCT dataset.
Experiment 1
Run a tracker on all videos in the NLPR_MCT dataset by using the single camera object tracking results in the ground truth annotation files, in other words, the identity relationships across cameras are not available. The available ground truth in the experiment is illustrated in Figure 1.
Experiment 2
Run a tracker on all videos in the NLPR_MCT dataset by using the ground truth bounding boxes. Any relationship between bounding boxes (i.e. identities) should not be utilized. The available ground truth in the experiment is illustrated in Figure 2.
Experiment 3
Run a tracker on all videos in the NLPR_MCT dataset without using any ground truth. The available ground truth in the experiment is illustrated in Figure 3.