基于级联注意力机制的孪生网络视觉跟踪算法

蒲磊; 冯新喜; 侯志强; 余旺盛; 马素刚

doi:10.13700/j.bh.1001-5965.2019.0601

基于级联注意力机制的孪生网络视觉跟踪算法

doi: 10.13700/j.bh.1001-5965.2019.0601

1.
空军工程大学研究生院, 西安 710077
2.
空军工程大学信息与导航学院, 西安 710077
3.
西安邮电大学计算机学院, 西安 710121

基金项目:

国家自然科学基金 61571458

国家自然科学基金 61703423

详细信息

作者简介:
蒲磊  男, 博士研究生。主要研究方向:目标跟踪

冯新喜  男, 博士, 教授。主要研究方向:信息融合

侯志强  男, 博士, 教授。主要研究方向:计算机视觉

余旺盛  男, 博士, 讲师。主要研究方向:模式识别

马素刚  男, 博士, 副教授。主要研究方向:目标跟踪

通讯作者:
侯志强, E-mail: hzq@xupt.edu.cn

中图分类号: TP391.4
计量
- 文章访问数: 633
- HTML全文浏览量: 65
- PDF下载量: 183
- 被引次数: 0
出版历程
- 收稿日期: 2019-11-25
- 录用日期: 2020-03-27
- 网络出版日期: 2020-12-20

Siamese network visual tracking algorithm based on cascaded attention mechanism

1.
Graduate College, Air Force Engineering University, Xi'an 710077, China
2.
Institute of Information and Navigation, Air Force Engineering University, Xi'an 710077, China
3.
School of Computer Science and Technology, Xi'an University of Posts and Telecommunications, Xi'an 710121, China

Funds:

National Natural Science Foundation of China 61571458

National Natural Science Foundation of China 61703423

More Information

Corresponding author: HOU Zhiqiang, E-mail: hzq@xupt.edu.cn

摘要

摘要:
针对全卷积孪生网络（SiamFC）在相似物体干扰及目标发生大尺度外观变化时容易跟踪失败的问题，提出了一种基于级联注意力机制的孪生网络视觉跟踪算法。首先，在网络的最后一层加入非局部注意力模块，从空间维度得到关于目标区域的自注意特征图，并与最后一层特征进行相加运算。其次，考虑到不同通道特征对不同目标和各类场景的响应差异，引入通道注意力模块实现对特征通道的重要性选择。为了进一步提高跟踪的鲁棒性，将其与SiamFC算法进行加权融合，得到最终的响应图。最后，将提出的孪生网络模型在GOT10k和VID数据集上进行联合训练，进一步提升模型的表达力与判别力。实验结果表明：所提算法相比于SiamFC，在跟踪精度上提高了9.3%，在成功率上提高了5.4%。
- 视觉跟踪 /
- 孪生网络 /
- 非局部注意力 /
- 通道注意力 /
- 模型集成
Abstract:
Aimed at the problem that the Fully Convolutional Siamese Network (SiamFC) is easy to fail to track when it is disturbed by similar object or the target has large-scale appearance changes, this paper proposes a Siamese network visual tracking algorithm based on cascaded attention mechanism. First, the non-local attention module is added to the last layer of the network, and the self-attention feature map of the target area is obtained from the spatial dimension and is added with the last-layer feature. Then, considering the different responses of different channel features to different targets and scenes, the channel attention module is introduced to select the importance of feature channel. In order to further improve the robustness of tracking, it is weighted fused with SiamFC algorithm to obtain the final response map. Finally, the Siamese network model is proposed to jointly train on the GOT10k and VID data set to further improve the expression and discrimination of the model. Experimental results show that compared with SiamFC, the proposed algorithm improves the accuracy by 9.3% and the success rate by 5.4%.
- visual tracking /
- Siamese network /
- non-local attention /
- channel attention /
- model integration

HTML全文

图 1 本文算法框架

Figure 1. Framework of proposed algorithm

下载: 全尺寸图片幻灯片

图 2 非局部注意力模块

Figure 2. Non-local attention module

下载: 全尺寸图片幻灯片

图 3 通道注意力模块

Figure 3. Channel attention module

下载: 全尺寸图片幻灯片

图 4 定性分析

Figure 4. Qualitative analysis

下载: 全尺寸图片幻灯片

图 5 不同算法的跟踪精度曲线和成功率曲线

Figure 5. Curves of distance precision and success rate of different algorithms

下载: 全尺寸图片幻灯片

图 6 不同属性下算法的跟踪精度对比曲线

Figure 6. Tracking precision comparison curves of algorithm under different attributes

下载: 全尺寸图片幻灯片

图 7 不同属性下算法的跟踪成功率对比曲线

Figure 7. Tracking success rate comparison curves of algorithm under different attributes

下载: 全尺寸图片幻灯片

图 8 算法关键环节对跟踪性能影响对比实验

Figure 8. Comparison experiment of influence of key parts of algorithm on tracking performance

下载: 全尺寸图片幻灯片

图 9 跟踪失败情况

Figure 9. Tracking failures

下载: 全尺寸图片幻灯片

表 1 深度学习算法跟踪速度对比

Table 1. Comparison of tracking speed of deep learning algorithms

算法	本文	HCF	CFNet	DCFNet	SiamFC
跟踪速度/(帧·s^-1)	58	10.2	78.4	65.9	83.7

下载: 导出CSV

参考文献(26)

[1]	SMEULDERS A W M, CHU D M, CUCCHIARA R, et al.Visual tracking:An experimental survey[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 36(7):1442-1468.
[2]	BOLME D S, BEVERIDGE J R, DRAPER B A, et al.Visual object tracking using adaptive correlation filters[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway: IEEE Press, 2010: 2544-2550.
[3]	HENRIQUES J F, CASEIRO R, MARTINS P, et al.Exploiting the circulant structure of tracking-by-detection with kernels[C]//Proceedings of the European Conference on Computer Vision.Berlin: Springer, 2012: 702-715.
[4]	HENRIQUES J F, RUI C, 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.
[5]	DANELLJAN M, SHAHBAZ K F, FELSBERG M, et al.Adaptive color attributes for real-time visual tracking[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway: IEEE Press, 2014: 1090-1097.
[6]	GIRSHICK R, DONAHUE J, DARRELL T, et al.Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway: IEEE Press, 2014: 580-587.
[7]	LONG J, SHELHAMER E, DARRELL T.Fully convolutional networks for semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway: IEEE Press, 2015: 3431-3440.
[8]	RAWAT W, WANG Z.Deep convolutional neural networks for image classification:A comprehensive review[J].Neural Computation, 2017, 29(9):2352-2449.
[9]	NAM H, HAN B.Learning multi-domain convolutional neural networks for visual tracking[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway: IEEE Press, 2016: 4293-4302.
[10]	DANELLJAN M, HAGER G, KHAN S F, et al.Convolutional features for correlation filter based visual tracking[C]//Proceedings of the IEEE International Conference on Computer Vision Workshops.Piscataway: IEEE Press, 2015: 58-66.
[11]	DANELLJAN M, ROBINSON A, KHAN F S, et al.Beyond correlation filters: Learning continuous convolution operators for visual tracking[C]//Proceedings of the European Conference on Computer Vision.Berlin: Springer, 2016: 472-488.
[12]	BHAT G, JOHNANDER J, DANELLJAN M, et al.Unveiling the power of deep tracking[C]//Proceedings of the European Conference on Computer Vision.Berlin: Springer, 2018: 483-498.
[13]	DANELLJAN M, BHAT G, KHAN S F, et al.ECO: Efficient convolution operators for tracking[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway: IEEE Press, 2017: 6931-6939.
[14]	BERTINETTO L, VALMADRE J, HENRIQUES J F, et al.Fully convolutional siamese networks for object tracking[C]//Proceedings of the European Conference on Computer Vision.Berlin: Springer, 2016: 850-865.
[15]	LI B, YAN J Y, WU W, et al.High performance visual tracking with siamese region proposal network[C]//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Piscataway: IEEE Press, 2018: 8971-8980.
[16]	GUO Q, FENG W, ZHOU C, et al.Learning dynamic siamese network for visual object tracking[C]//Proceedings of the IEEE International Conference on Computer Vision.Piscataway: IEEE Press, 2017: 1781-1789.
[17]	WU Y, LIM J, YANG M H.Object tracking benchmark[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(9):1834-1848.
[18]	WANG X, GIRSHICK R, GUPTA A, et al.Non-local neural networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway: IEEE Press, 2018: 7794-7803.
[19]	HU J, SHEN L, SUN G.Squeeze-and-excitation networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway: IEEE Press, 2018: 7132-7141.
[20]	MA C, HUANG J B, YANG X K, et al.Hierarchical convolutional features for visual tracking[C]//IEEE International Conference on Computer Vision.Piscataway: IEEE Press, 2015: 3074-3082.
[21]	BERTINETTO L, VALMADRE J, GOLODETZ S, et al.Staple: Complementary learners for real-time tracking[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway: IEEE Press, 2016: 1401-1409.
[22]	LI Y, ZHU J.A scale adaptive kernel correlation filter tracker with feature integration[C]//Proceedings of the European Conference on Computer Vision.Berlin: Springer, 2014: 254-265.
[23]	MA C, YANG X, ZHANG C, et al.Long-term correlation tracking[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway: IEEE Press, 2015: 5388-5396.
[24]	VALMADRE J, BERTINETTO L, HENRIQUES J, et al.End-to-end representation learning for correlation filter based tracking[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway: IEEE Press, 2017: 5000-5008.
[25]	WANG Q, GAO J, XING J L, et al.DCFNet: Discriminant correlation filters network for visual tracking[EB/OL].(2017-04-13)[2019-11-20].http://arxiv.org/abs/1704.04057.
[26]	ZHANG J, MA S, SCLAROFF S.MEEM: Robust tracking via multiple experts using entropy minimization[C]//Proceedings of the European Conference on Computer Vision.Berlin: Springer, 2014: 188-203.