Extending Multi-Object Detection Ability using Correlative Filter

Recently, numerous object detection methods are proposed and has been published by many papers and journals. Since the big hit of YOLO object detector in 2014, even lighter (thus faster) or more accurate method has been developed with good success. On the other side of computer vision advancement, the annual Visual Object Tracker (VOT) challenge is presenting us the newest state-of-the-art tracking methods, which are getting better for each year both in terms of accuracy and speed. In an object detection system with a goal to identify the surrounding environment, it is beneficial to take advantages from rapid development in both object detection and object tracking method. The combined method is able to automatically detect and identify which object is which in subsequent frames with a considerable performance for indoor settings.