May I only use ByteTrack to improve the result of detection ?

[le02146]

I am using yolo model to run object detect, instead of tracking. I trained yolo on many images with bboxes of car, then run car detection on a video/images. The problem is that some of the frames are missing which has no any predicted detection, may I use ByteTrack method to improve(postprocessing or offline or online?) the result of detection? For example, could ByteTrack predict the bbox of one car when yolo could not output any car bbox in a frame. Thanks.