Deep Deblurring Correlation Filter for Object Tracking

TY - JOUR

T1 - Deep Deblurring Correlation Filter for Object Tracking

AU - Bai, Yu

AU - Xu, Tingfa

AU - Huang, Bo

AU - Yang, Ruoling

N1 - Publisher Copyright: © 2013 IEEE.

PY - 2020

Y1 - 2020

N2 - Motion blur is a quite tricky issue in object tracking community. In recent years, many trackers have been explored to address this issue without sensational performance. This paper proposes a novel correlation tracking framework with Recurrent Neural Network (RNN) deblurnet and proposal detection to handle motion blur and heavy occlusion in object tracking tasks. We take advantage of high efficient Kernelized Correlation Filter (KCF) tracker, a typical method that exploits the circulant structure and the kernel trick to enhance the performance, and furthermore incorporate two regression methods in it. We employ RNN as our baseline of deblurnet, and introduce residual block and ConvLSTM in our deblur network to improve the result of deblurring. In addition, we suggest an edge information based rectification system to overcome the challenge of target occlusion. Finally, we update the model adaptively in term of the feedback from high-confidence tracking results to avoid the model degradation. Extensive experimental results demonstrate our tracker outperforms several state-of-the-art trackers on the OTB-2015 and VOT-2016 datasets.

AB - Motion blur is a quite tricky issue in object tracking community. In recent years, many trackers have been explored to address this issue without sensational performance. This paper proposes a novel correlation tracking framework with Recurrent Neural Network (RNN) deblurnet and proposal detection to handle motion blur and heavy occlusion in object tracking tasks. We take advantage of high efficient Kernelized Correlation Filter (KCF) tracker, a typical method that exploits the circulant structure and the kernel trick to enhance the performance, and furthermore incorporate two regression methods in it. We employ RNN as our baseline of deblurnet, and introduce residual block and ConvLSTM in our deblur network to improve the result of deblurring. In addition, we suggest an edge information based rectification system to overcome the challenge of target occlusion. Finally, we update the model adaptively in term of the feedback from high-confidence tracking results to avoid the model degradation. Extensive experimental results demonstrate our tracker outperforms several state-of-the-art trackers on the OTB-2015 and VOT-2016 datasets.

KW - Object tracking

KW - RNN-deblurnet

KW - adaptive update

KW - correlation filter

KW - motion blur

UR - https://www.scopus.com/pages/publications/85084188685

U2 - 10.1109/ACCESS.2020.2986311

DO - 10.1109/ACCESS.2020.2986311

M3 - Article

AN - SCOPUS:85084188685

SN - 2169-3536

VL - 8

SP - 68623

EP - 68637

JO - IEEE Access

JF - IEEE Access

M1 - 9058688

ER -