Comparative Evaluations of Selected Tracking-by-Detection Approaches
Résumé
In this work, we present a comparative evaluation of various multi-person tracking-by-detection approaches on public datasets. The work investigates five popular trackers coupled with six relevant visual people detectors evaluated on seven public datasets. The evaluation emphasizes on exhibited performance variation depending on tracker-detector choices. Our experimental results show that the overall performance depends on how challenging the dataset is, the performance of the detector on the specific dataset, and the tracker-detector combination. Some trackers are more sensitive to the choice of a detector and some detectors to the choice of a tracker than others. Based on our results, two of the trackers demonstrate the best performances consistently across different datasets whereas the best performing detectors vary per dataset. This underscores the need for careful application context specific evaluation when choosing a detector.
Origine | Fichiers éditeurs autorisés sur une archive ouverte |
---|
Loading...