Research on the Frameworks of Visual Tracking

With the development of visual tracking, the visual tracking plays a significant role in various fileds: traffic control, surveillance, facial recognition and so on. In this paper, I decide to introduce three kinds of framework about the Transformer Tracking, MDNet and correlation filter. At the same time, I will find out the characteristic of these three frameworks. By comparing each of these frameworks, I will analyze the advantages and disadvantages of these frameworks. I also provide the in-depth analysis on each framework and provide a brief guidance on how to select them. We note that our paper can provide an introduction of visual object tracking for those beginners.