Deep Tracking With Objectness

 Xinyu Wang¹      Hanxi Li¹*      Yi Li²      Fumin Shen³      Fatih Porikli⁴      Mingwen Wang¹

Jiangxi Normal University, China¹

Toyota Research Institute of North America, USA²

University of Electronic Science and Technology of China, China³

Australian National University, Australia⁴

1. Abstract

Visual tracking is a fundamental problem in computer vision. However, due to the (sometimes) ambiguous target information given at the first frame, it has also been criticized as less well-posed compared with other tasks with clearly defined targets, such as object detection and semantic segmentation. In this paper, we try to evaluate the importance of object category in visual tracking by tracking objects with known object types. The proposed algorithm, termed Deep Track with Objectness (DTO), naturally combines the state-of-the-art deep-learning-based detectors and trackers, which essentially share a large part of the network. In DTO, a deep tracker, which is scale-fixed and sensitive to small translations tracks the object in a relative short lifespan. A deep detector, which is scale-changeable and robust to pose or illumination changes guides the deep tracker in a longer lifespan. As the deep tracker and detector share the main part of their networks, no much extra computation is imposed while the performance gain is significant. We test the proposed algorithm on two well-accepted benchmarks and on both of them, the proposed method increases the tracking accuracies remarkably compared with state-of-the-art visual trackers.

 

2. Downloads

Deep Tracking With Objectness

Xinyu Wang, Hanxi Li*, Yi Li, Fumin Shen, Fatih Porikli, Mingwen Wang

IEEE International Conference on Image Processing (ICIP) 2017, Beijing (Oral)

[Paper]     [Code Coming Soon]

[Results Coming Soon]

 

3. Evaluate Results

Location Error of OTB50 Car Subset
Success Rate of OTB50 Car Subset

 

4. Conclusion and Future Works

In this paper, we propose a very simple yet effective way to guide the visual tracking by the detection results. The proposed DTO tracker can be considered as a fusion of the state-of-the-art deep tracker and deep detector. As they share most part of the network structure, no much extra computation is required. On the other hand, we can see a dramatic performance improvement in DTO, compared with its prototype, the HCF tracker. This improvement implies the absence of the target object could lead to poor tracking performance while to address this absence in a more sophisticated way could yield much better deep trackers in the future.

 

5. Reference

If you feel this research is helpful, please consider cite our paper.

@inproceedings{wang2017deep,
  title={Deep Tracking With Objectness},
  author={Wang, Xinyu and Li, Hanxi and Li, Yi and Shen, Fumin and Porikli, Fatih and Wang, Mingwen},
  booktitle={Proceeding of the IEEE International Conference on Image and Processing (ICIP)},
  year={2017}
}