Journal of Frontiers of Computer Science and Technology ›› 2022, Vol. 16 ›› Issue (6): 1405-1416.DOI: 10.3778/j.issn.1673-9418.2012016

• Graphics and Image • Previous Articles     Next Articles

High Frame Rate Light-Weight Siamese Network Target Tracking

LI Yunhuan, WEN Jiwei, PENG Li()   

  1. Engineering Research Center of Internet of Things Technology Applications (School of Internet of Things Engineering, Jiangnan University), Ministry of Education, Wuxi, Jiangsu 214122, China
  • Received:2020-12-03 Revised:2021-01-29 Online:2022-06-01 Published:2021-02-04
  • About author:LI Yunhuan, born in 1998, M.S. candidate. His research interests include deep learning, computer vision and target tracking.
    WEN Jiwei, born in 1981, Ph.D., associate professor, M.S. supervisor. His research interests include stochastic switched systems, model predictive control, T-S fuzzy modeling and control.
    PENG Li, born in 1967, Ph.D., professor, Ph.D. supervisor, member of CAAI and CCF. His research interests include visual Internet of things, action recognition and deep learning.
  • Supported by:
    National Key Research and Development Program of China(2018YFD0400902);National Natural Science Foundation of China(61873112)


李运寰, 闻继伟, 彭力()   

  1. 物联网技术应用教育部工程研究中心(江南大学 物联网工程学院),江苏 无锡 214122
  • 通讯作者: + E-mail:
  • 作者简介:李运寰(1998—),男,江苏盐城人,硕士研究生,主要研究方向为深度学习、计算机视觉、目标跟踪。
    彭力(1967—),男,河北唐山人,博士,教授,博士生导师,CAAI会员,CCF会员, 主要研究方向为视觉物联网、行为识别、深度学习。
  • 基金资助:


With the widespread use of target tracking in many life scenarios, the demand for high-precision and high-speed tracking algorithms is also increasing. For some specific scenarios such as mobile terminals, embedded devices, etc., under the premise of relatively insufficient computing power of the device, it is still necessary to ensure that the tracker achieves good tracking accuracy and high-speed real-time tracking. A high frame rate tracking algorithm based on light-weight siamese network is proposed to solve this problem. Firstly, the light-weight convolutional neural network MobileNetV1 is selected, which is easy to be deployed in embedded devices, as the feature extraction backbone network, and deep network is more capable of extracting target features. Then, two optimization strategies are proposed to address the shortcomings of the backbone network, feature map is cropped and the total network step length is adjusted to make the backbone network suitable for tracking tasks. Finally, after the template branch of the siamese network, an ultra-lightweight channel attention module is added to weight important information that highlights the target characteristics. The proposed algorithm parameters are reduced by 59.8% in comparison with current mainstream algorithm SiamFC. Simulation and experimental results on the OTB2015 dataset show that the tracking accuracy is increased by 5.4%, and the algorithm can better cope with complex and changeable challenges in tracking tasks. Simulation and experimental results on the VOT2018 dataset show that the comprehensive index expected average overlap (EAO) is increased by 26.6%, and the average speed of the algorithm under NVIDIA GTX1080Ti is 120 frame/s, which achieves high frame rate real-time tracking.

Key words: target tracking, MobileNet, siamese networks, channel attention mechanism


随着目标跟踪在众多生活场景的广泛运用,高精度且高速的跟踪算法需求也日益增多。针对某些特定场景如移动端、嵌入式等设备,在设备算力相对不足的前提下,仍要保证跟踪器达到良好的跟踪精度和高速实时跟踪问题,提出一种高帧率的轻量级孪生网络目标跟踪算法。首先,选取易于部署在嵌入式设备中的轻量级卷积神经网络MobileNetV1作为特征提取网络,深层网络具有对目标特征强大的提取能力;接着,针对主干网络的不足提出两点优化策略,特征图裁剪和网络总步长调整,使得主干网络适用于跟踪任务;最后,在孪生网络的模板分支后添加超轻量级通道注意力模块,加权突出目标特征的重要信息。对比当前主流算法SiamFC,该算法参数量减少59.8%;在OTB2015数据集上仿真实验表明,跟踪精度提升了5.4%,算法能更好地应对跟踪任务中复杂多变的挑战;在VOT2018数据集上的仿真实验表明,综合指标平均重叠期望(EAO)提升了26.6%,同时算法在NVIDIA GTX1080Ti下的平均速度为120 frame/s,达到高帧率实时跟踪。

关键词: 目标跟踪, MobileNet, 孪生网络, 通道注意力机制

CLC Number: