[1]JIANG Wentao,MENG Qingjiao.Correlation filter tracking for adaptive spatiotemporal regularization[J].CAAI Transactions on Intelligent Systems,2023,18(4):754-763.[doi:10.11992/tis.202202030]
Copy

Correlation filter tracking for adaptive spatiotemporal regularization

CAAI Transactions on Intelligent Systems[ISSN 1673-4785/CN 23-1538/TP] Volume: 18 Number of periods: 2023 4 Page number: 754-763 Column: 学术论文—机器感知与模式识别 Public date: 2023-07-15
Title:
Correlation filter tracking for adaptive spatiotemporal regularization
Author(s):
JIANG Wentao;  MENG Qingjiao
School of Software, Liaoning Technical University, Huludao 125105, China
Keywords:
spatiotemporal adaptation; local response; global response; neural network; convolutional neural network; feature extraction; dimension reduction; principal component analysis algorithm
CLC:
TP391.4
DOI:
10.11992/tis.202202030
Abstract:
The regularization filter defines the regularization term in advance, but it can not suppress the learning of non-target region in real time. Aiming at this shortcoming, a new adaptive spatiotemporal regularization method is proposed to improve robustness of the algorithm to adapt to appearance change in the process of target tracking. The method firstly introduces the spatial local response variation into the objective function, so that the filter can focus on the trusted part of the learning object, obtaining the response model. Then, the update rate of the filter is determined based on the change of global response. Finally, the convolutional neural network is introduced for deep feature extraction. In order to reduce the storage capacity of high-dimensional data, the principal component analysis algorithm is employed to reduce dimension, which not only retains the main features, but also speeds up the calculation. Comparing with the spatial-temporal regularized correlation filtering algorithms, the average accuracy and success rate of data sets OTB2013 and OTB2015 are respectively increased by 4.7% and 12.7%. A large number of experiments show that the algorithm basically meets the real-time requirements in complex background, object occlusion, fast motion and other scenes.
References:
[1] 李馨, 赵加清, 张征明, 等. 基于数字图像相关的自适应应变场计算[J]. 光学学报, 2021, 41(23): 125–132
LI Xin, ZHAO Jiaqing, ZHANG Zhengming, et al. Self-adaptive strain field calculation based on digital image correlation[J]. Acta optica sinica, 2021, 41(23): 125–132
[2] 刘万军, 孙虎, 姜文涛. 自适应特征选择的相关滤波跟踪算法[J]. 光学学报, 2019, 39(6): 242–255
LIU Wanjun, SUN Hu, JIANG Wentao. Correlation filter tracking algorithm for adaptive feature selection[J]. Acta optica sinica, 2019, 39(6): 242–255
[3] 孙德刚, 肖媛媛, 尹艳华, 等. 多特征联合时空正则化相关滤波目标跟踪鲁棒算法[J]. 机床与液压, 2021, 49(22): 67–75
SUN Degang, XIAO Yuanyuan, YIN Yanhua, et al. Correlation filtering target tracking algorithm with multi-feature joint spatio-temporal regularization[J]. Machine tool & hydraulics, 2021, 49(22): 67–75
[4] BOLME D S, BEVERIDGE J R, DRAPER B A, et al. Visual object tracking using adaptive correlation filters[C]//2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. San Francisco: IEEE, 2010: 2544?2550.
[5] HENRIQUES J F, CASEIRO R, MARTINS P, et al. Exploiting the circulant structure of tracking-by-detection with kernels[C]//European Conference on Computer Vision. Berlin: Springer, 2012: 702?715.
[6] HENRIQUES J F, CASEIRO R, MARTINS P, et al. High-speed tracking with kernelized correlation filters[J]. IEEE transactions on pattern analysis and machine intelligence, 2015, 37(3): 583–596.
[7] DANELLJAN M, H?GER G, SHAHBAZ KHAN F, et al. Accurate scale estimation for robust visual tracking[C]// Proceedings of the British Machine Vision Conference. Nottingham: BMVA Press, 2014.
[8] BERTINETTO L, VALMADRE J, GOLODETZ S, et al. Staple: complementary learners for real-time tracking[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 1401?1409.
[9] GALOOGAHI H K, SIM T, LUCEY S. Correlation filters with limited boundaries[C]//2015 IEEE Conference on Computer Vision and Pattern Recognition. Boston: IEEE, 2015: 4630?4638.
[10] DANELLJAN M, H?GER G, KHAN F S, et al. Convolutional features for correlation filter based visual tracking[C]//2015 IEEE International Conference on Computer Vision Workshop. Santiago: IEEE, 2016: 621?629.
[11] MUELLER M, SMITH N, GHANEM B. Context-aware correlation filter tracking[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017: 1387?1395.
[12] LI Feng, TIANG Cheng, ZUO Wangmeng, et al. Learning spatial-temporal regularized correlation filters for visual tracking[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 4904–4913.
[13] HUANG Ziyuan, FU Changhong, LI Yiming, et al. Learning aberrance repressed correlation filters for real-time UAV tracking[C]//2019 IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2020: 2891?2900.
[14] 韦联福. 再谈拉格朗日函数非唯一性的物理意义: 广义规范变换和正则变换的等价性证明[J]. 物理与工程, 2021, 31(2): 31–35,40
WEI Lianfu. Remarks on the non-uniqueness of Lagrange function: an equivalence proof between the gauge and canonical transformations[J]. Physics and engineering, 2021, 31(2): 31–35,40
[15] 姜文涛, 刘晓璇, 涂潮, 等. 自适应空间异常的目标跟踪[J]. 电子与信息学报, 2022, 44(2): 523–533
JIANG Wentao, LIU Xiaoxuan, TU Chao, et al. Adaptive spatial and anomaly target tracking[J]. Journal of electronics & information technology, 2022, 44(2): 523–533
[16] 孙丽君, 黄志远, 陈天飞. 基于傅里叶变换的Gamma因子快速自标定方法[J]. 光学学报, 2021, 41(24): 116–127
SUN Lijun, HUANG Zhiyuan, CHEN Tianfei. Fast self-calibration method of gamma factor based on Fourier transform[J]. Acta optica sinica, 2021, 41(24): 116–127
[17] 李文健, 盖绍彦, 俞健, 等. 基于卷积神经网络的单帧复合图像绝对相位恢复[J]. 光学学报, 2021, 41(23): 113–124
LI Wenjian, GAI Shaoyan, YU Jian, et al. Absolute phase recovery of single frame composite image based on convolutional neural network[J]. Acta optica sinica, 2021, 41(23): 113–124
[18] 高宇, 夏志明, 刘欢. 非线性二维主成分分析方法[J]. 纯粹数学与应用数学, 2021, 37(4): 475–492
GAO Yu, XIA Zhiming, LIU Huan. Two-dimensional nonlinear principal component analysis: a nonlinear information compression method[J]. Pure and applied mathematics, 2021, 37(4): 475–492
[19] 梁霄, 李家炜, 赵小龙, 等. 基于深度学习的红外目标成像液位检测方法[J]. 光学学报, 2021, 41(21): 104–112
LIANG Xiao, LI Jiawei, ZHAO Xiaolong, et al. Infrared target imaging liquid level detection method based on deep learning[J]. Acta optica sinica, 2021, 41(21): 104–112
[20] 刘宗达, 董立泉, 赵跃进, 等. 视频中快速运动目标的自适应模型跟踪算法[J]. 光学学报, 2021, 41(18): 164–173
LIU Zongda, DONG Liquan, ZHAO Yuejin, et al. Adaptive model tracking algorithm for fast-moving targets in video[J]. Acta optica sinica, 2021, 41(18): 164–173
[21] 李天雄, 侯茂盛, 李丽娟, 等. 基于特征自适应的激光扫描投影图形控制点提取及优化方法[J]. 光学学报, 2021, 41(24): 98–109
LI Tianxiong, HOU Maosheng, LI Lijuan, et al. Control point extraction and optimization method of laser scanning projection graphics based on feature adaptation[J]. Acta optica sinica, 2021, 41(24): 98–109
[22] 尹宝才, 张超辉, 胡永利, 等. 基于图嵌入的自适应多视降维方法[J]. 智能系统学报, 2021, 16(5): 963–970
YIN Baocai, ZHANG Chaohui, HU Yongli, et al. An adaptive multi-view dimensionality reduction method based on graph embedding[J]. CAAI transactions on intelligent systems, 2021, 16(5): 963–970
[23] 王德文, 魏波涛. 基于孪生变分自编码器的小样本图像分类方法[J]. 智能系统学报, 2021, 16(2): 254–262
WANG Dewen, WEI Botao. A small-sample image classification method based on a Siamese variational auto-encoder[J]. CAAI transactions on intelligent systems, 2021, 16(2): 254–262
[24] 崔铁军, 李莎莎. 基于因素空间的人工智能样本选择策略[J]. 智能系统学报, 2021, 16(2): 346–352
CUI Tiejun, LI Shasha. Sample selection strategy of artificial intelligence based on factor space[J]. CAAI transactions on intelligent systems, 2021, 16(2): 346–352
[25] 石昌友, 孙强, 卢建平, 等. 基于深度融合卷积神经网络的图像边缘检测[J]. 现代电子技术, 2022, 45(24): 141–144
SHI Chagyou, SUN Qiang, LU Jianping, et al. Image edge detection based on deep fusion convolution neural network[J]. Modern electronic technology, 2022, 45(24): 141–144
Similar References:

Memo

-

Last Update: 1900-01-01

Copyright © CAAI Transactions on Intelligent Systems