Abstract
We investigate the application of ByteTrack in the realm of multiple objecttracking. ByteTrack, a simple tracking algorithm, enables the simultaneoustracking of multiple objects by strategically incorporating detections with alow confidence threshold. Conventionally, objects are initially associated withhigh confidence threshold detections. When the association between objects anddetections becomes ambiguous, ByteTrack extends the association to lowerconfidence threshold detections. One notable drawback of the existing ByteTrackapproach is its reliance on a fixed threshold to differentiate between high andlow-confidence detections. In response to this limitation, we introduce a noveland adaptive approach. Our proposed method entails a dynamic adjustment of theconfidence threshold, leveraging insights derived from overall detections.Through experimentation, we demonstrate the effectiveness of our adaptiveconfidence threshold technique while maintaining running time compared toByteTrack.