Skip to main content

Advertisement

Springer Nature Link
Log in

Learning deep convolutional descriptor aggregation for efficient visual tracking

  • Original Article
  • Published:
Neural Computing and Applications Aims and scope Submit manuscript

Abstract

Visual trackers have achieved a high-level performance from deep features, but many limitations remain. Online trackers suffer from low speed while using deep features for parameter updating, and deep trackers trained offline demonstrate data hunger. To meet these challenges, our work aims to mine the target representation capability of a pre-trained model and presents deep convolutional descriptor aggregation (DCDA) for visual tracking. Based on spatial and semantic priors, we propose an edge-aware selection (EAS) and a central-aware selection (CAS) method to aggregate the accuracy-aware and robustness-aware features. To make full use of the scene context, our method is derived from one-shot learning by designing a dedicated regression process that is capable of predicting discriminative model in a few iterations. By exploiting robustness feature aggregation, the accuracy feature aggregation, and the discriminative regression, our DCDA with Siamese tracking architecture not only enhances the target prediction capacity, but also achieves a low-cost reuse of the pre-trained model. Comprehensive experiments on OTB-100, VOT2016, VOT2017, VOT2020, NFS30, and NFS240 show that our DCDA tracker achieves state-of-the-art performance with a high running speed of 65 FPS. The source code and all the experimental results of this work will be made public at https://github.com/Gitlyz007/DCDA_Tracker.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Subscribe and save

Springer+ Basic
$34.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14
Fig. 15
Fig. 16
Fig. 17

Similar content being viewed by others

Explore related subjects

Discover the latest articles and news from researchers in related subjects, suggested using machine learning.

Notes

  1. The pre-trained models are obtained from: https://www.vlfeat.org/matconvnet/pretrained/.

References

  1. Bau D, Zhou B, Khosla A, Oliva A, Torralba A(2017) Network dissection: quantifying interpretability of deep visual representations. In: Proceedings of the CVPR, pp 6541–6549

  2. Bertinetto L, Henriques J, Valmadre J, Torr P, Vedaldi A (2016) Learning feed-forward one-shot learners. In: Proceeding of the NIPS, pp 523–531

  3. Bertinetto L, Valmadre J, Golodetz S, Miksik O, Torr PHS (2016) Staple: complementary learners for real-time tracking. In: Proceeding of the CVPR, pp 1401–1409

  4. Bertinetto L, Valmadre J, Henriques JF, Vedaldi A, Torr PHS (2016) Fully-convolutional siamese networks for object tracking. In: Proceedings of the ECCVW, pp 850–865. Springer

  5. Zhizhen C, Hongyang L, Huchuan L, Ming-Hsuan Y (2017) Dual deep network for visual tracking. IEEE Trans Image Process 26(4):2005–2015

    Article  MathSciNet  Google Scholar 

  6. Choi J, Jin CH, Fischer T, Yun S, Lee K, Jeong J, Demiris Y, Young CJ (2018) Context-aware deep feature compression for high-speed visual tracking. In: Proceeding of the CVPR, pp 479–488

  7. Chu P, Ling H (2019) Famnet: Joint learning of feature, affinity and multi-dimensional assignment for online multiple object tracking. In: Proceeding of the CVPR, pp 6172–6181

  8. Danelljan M (2018) Learning convolution operators for visual tracking, vol 1926. Linköping University Electronic Press, Linköping

  9. Danelljan M, Häger G, Khan FS, Felsberg M (2015) Coloring channel representations for visual tracking. In: Scandinavian conference on image analysis, pp 117–129. Springer

  10. Danelljan M, Hager G, Shahbaz KF, Felsberg M (2015) Convolutional features for correlation filter based visual tracking. In: Proceeding of the ICCVW, pp 58–66

  11. Danelljan M, Hager G, Shahbaz KF, Felsberg M (2015) Learning spatially regularized correlation filters for visual tracking. In: Proceeding of the ICCV, pp 4310–4318

  12. Martin D, Gustav H, Shahbaz KF, Michael F (2016a) Discriminative scale space tracking. IEEE Trans Pattern Anal Mach Intell 39(8):1561–1575

    Google Scholar 

  13. Danelljan M, Robinson A, Khan FS, Felsberg M (2016) Beyond correlation filters: learning continuous convolution operators for visual tracking. In: Proceeding of the ECCV, pp. 472–488. Springer

  14. Danelljan M, Bhat G, Shahbaz KF, Felsberg M (2017) Eco: efficient convolution operators for tracking. In: Proceeding of the CVPR, pp 6638–6646

  15. Dong X, Shen J (2018) Triplet loss in siamese network for object tracking. In: Proceeding of the ECCV, pp 459–474

  16. Fan H, Lin L, Yang F, Chu P, Deng G, Yu S, Bai H, Xu Y, Liao Y, Ling Y (2019) Lasot: a high-quality benchmark for large-scale single object tracking. In: Proceeding of the CVPR, pp 5374–5383

  17. Finn C, Abbeel P, Levine S (2017) Model-agnostic meta-learning for fast adaptation of deep networks. In: Proceedings of the ICML, pp 1126–1135. JMLR. org

  18. Gao J, Zhang T, Xu C (2019) Graph convolutional tracking. In: Proceedings of the CVPR, pp 4649–4659

  19. He A, Luo C, Tian X, Zeng W (2018) A twofold siamese network for real-time object tracking. In: Proceedings of the CVPR, pp 4834–4843

  20. He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. In: Proceedings of the CVPR, pp 770–778

  21. He Z, Fan Y, Zhuang J, Dong Y, Bai HL (2017) Correlation filters with weighted convolution responses. In: Proceedings of the ICCVW, pp 1992–2000

  22. Held D, Thrun S, Sav S (2016) Learning to track at 100 fps with deep regression networks. In: Proceedings of the ECCV, pp 749–765. Springer

  23. Henriques João F, Rui C, Pedro M, Jorge B (2014) High-speed tracking with kernelized correlation filters. IEEE Trans Pattern Anal Mach Intell 37(3):583–596

    Article  Google Scholar 

  24. Kiani GH, Sim T, Lucey S (2015) Correlation filters with limited boundaries. In: Proceedings of the CVPR, pp 4630–4638

  25. Kiani GH, Fagg A, Huang C, Ramanan D, Lucey S (2017) Need for speed: a benchmark for higher frame rate object tracking. In: Proceedings of the ICCV, pp 1125–1134

  26. Kiani GH, Fagg A, Lucey S (2017) Learning background-aware correlation filters for visual tracking. In: Proceedings of the ICCV, pp 1135–1143

  27. Kristan M, Lukezic A, Danelljan M, Čehovin ZL, Matas J (2020) The new vot2020 short-term tracking performance evaluation protocol and measures

  28. Kristan M, Matas J, Leonardis A, Felsberg M, Cehovin L, Fernández G, Vojir H, Tomas et al (2016) The visual object tracking vot2016 challenge results. In: Proceedings of the ECCVW, vol 2, p 8

  29. Kristan M, Matas J, Leonardis A, Vojir T, Pflugfelder R, Fernandez G, Nebehay G, Porikli F, Čehovin L (2016) A novel performance evaluation methodology for single-target trackers. IEEE Trans Pattern Anal Mach Intell 38(11):2137–2155. https://doi.org/10.1109/TPAMI.2016.2516982

    Article  Google Scholar 

  30. Kristan M, Leonardis A, Matas A, Felsberg M, Pflugfelder R, Cehovin ZL, Vojir L, Hager G, Lukezic A, Eldesokey A et al (2017) The visual object tracking vot2017 challenge results. In: Proceedings of the ICCVW, pp 1949–1972

  31. Matej K, Jiri M, Ales L, Michael F, Roman P, Joni-Kristian K, Luka CZ, Ondrej D, Alan L, Amanda B et al (2019) The seventh visual object tracking vot2019 challenge results. In: Proceedings of the ICCVW

  32. Li B, Yan J, Wu W, Zhu Z, Hu X (2018) High performance visual tracking with siamese region proposal network. In: Proceedings of the CVPR, pp 8971–8980

  33. Li B, Wu W, Wang Q, Zhang F, Xing F, Yan J (2019) Siamrpn++: evolution of siamese visual tracking with very deep networks. In: Proceedings of the CVPR, pp 4282–4291

  34. Li P, Chen B, Ouyang W, Wang D, Yang X, Lu X (2019) Gradnet: gradient-guided network for visual object tracking. In: Proceedings of the ICCV, pp 6162–6171

  35. Li X, Ma C, Wu B, He Z, Yang MH (2019) Target-aware deep tracking. In: Proceedings of the CVPR, pp 1369–1378

  36. Li Y, Zhu J (2014) A scale adaptive kernel correlation filter tracker with feature integration. In: Proceedings of the ECCV, pp 254–265. Springer

  37. Yang L, Jianke Z, Hoi Steven CH, Wenjie S, Zhefeng W, Hantang L (2019) Robust estimation of similarity transformation for visual object tracking. In: Proc AAAI 33:8666–8673

  38. Shuai L, Shuai W, Xinyu L, Chin-Teng L, Zhihan L (2020) Fuzzy detection aided real-time and robust visual tracking under complex environments. IEEE Trans Fuzzy Syst

  39. Shuai L, Xinyu L, Shuai W, Khan M (2021) Fuzzy-aided solution for out-of-view challenge in visual tracking under iot-assisted complex environment. Neural Comput Appl 33:1055–1065

    Article  Google Scholar 

  40. Shuai L, Shuai W, Xinyu L, Gandomi Amir H, Mahmoud D, Khan M, de Albuquerque Victor Hugo C, (2021) Human memory update strategy: a multi-layer template update mechanism for remote visual monitoring. IEEE Trans Multimedia

  41. Wenxi L, Yibing S, Dengsheng C, He Shengfeng Yu, Yuanlong YT, Hancke Gehard P, Lau Rynson WH (2019) Deformable object tracking with gated fusion. IEEE Trans Image Process 28(8):3766–3777

    Article  MathSciNet  Google Scholar 

  42. Ma C, Yang X, Zhang C, Yang MH (2015) Long-term correlation tracking. In: Proceedings of the CVPR, pp 5388–5396

  43. Chao M, Jia-Bin H, Xiaokang Y, Ming-Hsuan Y (2018) Robust visual tracking via hierarchical convolutional features. IEEE Trans Pattern Anal Mach Intell 41(11):2709–2723

    Google Scholar 

  44. Marvasti-Zadeh MH, Ghanei-Yakhdan H, Kasaei S (2021) Efficient scale estimation methods using lightweight deep convolutional neural networks for visual tracking. Neural Comput Appl, pp 1–16

  45. Munkhdalai T, Yu H (2017) Meta networks. In: Proceedings of the ICML, pp 2554–2563. JMLR. org

  46. Nam H, Han B (2016) Learning multi-domain convolutional neural networks for visual tracking. In: Proceedings of the CVPR, pp 4293–4302

  47. Zaiyu P, Jun W, Guoqing W, Jihong Z (2020) Multi-scale deep representation aggregation for vein recognition. IEEE Trans Inf Forens Security 16:1–15

    Google Scholar 

  48. Adam P, Sam G, Francisco M, Adam L, James B, Gregory C, Trevor K, Zeming L, Natalia G, Luca A et al (2019) Pytorch: An imperative style, high-performance deep learning library. In: Proceedings of the NIPS 8024–8035

  49. Yuankai Q, Shengping Z, Lei Q, Qingming H, Hongxun Y, Jongwoo L, Ming-Hsuan Y (2018) Hedging deep features for visual tracking. IEEE Trans Pattern Anal Mach Intell 41(5):1116–1130

    Google Scholar 

  50. Real E, Shlens J, Mazzocchi S, Pan X, Vanhoucke V (2017) Youtube-boundingboxes: a large high-precision human-annotated data set for object detection in video. In: Proceedings of the CVPR, pp 5296–5305

  51. Olga R, Jia D, Hao S, Jonathan K, Sanjeev S, Sean M, Zhiheng H, Andrej K, Aditya K, Michael B, Berg Alexander C, Li F-F (2015) Imagenet large scale visual recognition challenge. Int J Comput Vis (IJCV) 115(3):211–252. https://doi.org/10.1007/s11263-015-0816-y

    Article  MathSciNet  Google Scholar 

  52. Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556

  53. Sun C, Wang D, Lu H, Yang M-H (2018) Learning spatial-aware regressions for visual tracking. In: Proceedings of the CVPR, pp 8962–8970

  54. Szegedy C, Liu W, Jia Y, Sermanet P, Reed P, Anguelov D, Erhan D, Vanhoucke V, Rabinovich A (2015) Going deeper with convolutions. In: Proceedings of the CVPR, pp 1–9

  55. Valmadre J, Bertinetto L, Henriques J, Vedaldi A, Torr PHS (2017) End-to-end representation learning for correlation filter based tracking. In: Proceedings of the CVPR, pp 2805–2813

  56. Wang G, Luo C, Xiong Z, Zeng Z (2019) Spm-tracker: series-parallel matching for real-time visual object tracking. In: Proceedings of the CVPR, pp 3643–3652

  57. Wang L, Ouyang W, Wang X, Lu H (2015) Visual tracking with fully convolutional networks. In: Proceedings of the ICCV, pp 3119–3127

  58. Wang N, Song Y, Ma C, Zhou W, Liu W, Li H (2019) Unsupervised deep tracking. In: Proceedings of the CVPR, pp 1308–1317

  59. Xiu-Shen W, Jian-Hao L, Jianxin W, Zhi-Hua Z (2017) Selective convolutional descriptor aggregation for fine-grained image retrieval. IEEE Trans Image Process 26(6):2868–2881

    Article  MathSciNet  Google Scholar 

  60. Xiu-Shen W, Chen-Lin Z, Jianxin W, Chunhua S, Zhi-Hua Z (2019) Unsupervised object discovery and co-localization by deep descriptor transformation. Pattern Recogn 88:113–126

    Article  Google Scholar 

  61. Wu Y, Lim J, Yang M-H (2013) Online object tracking: a benchmark. In: Proceedings of the CVPR, pp 2411–2418

  62. Yi W, Jongwoo L, Ming-Hsuan Y (2015) Object tracking benchmark. IEEE Trans Pattern Anal Mach Intell 37(9):1834–1848

    Article  Google Scholar 

  63. Xu J, Shi C, Qi C, Wang C, Xiao B (2018) Unsupervised part-based weighting aggregation of deep convolutional features for image retrieval. In: Proceedings of the AAAI, vol 32

  64. Kang Y, Huihui S, Kaihua Z, Qingshan L (2020) Hierarchical attentive siamese network for real-time visual tracking. Neural Comput Appl 32(18):14335–14346

    Article  Google Scholar 

  65. Yang T, Chan AB (2018) Learning dynamic memory networks for object tracking. In: Proceedings of the ECCV, pp 152–167

  66. Tianyu Y, Chan Antoni B (2019) Visual tracking via dynamic memory networks. IEEE Trans Pattern Anal Mach Intell

  67. Yang Y, De-Chuan Z, Ying F, Yuan J, Zhi-Hua Z (2017) Deep learning for fixed model reuse. In: Proceedings of the AAAI

  68. Yin J, Wang W, Meng Q, Yang R, Shen J (2020) A unified object motion and affinity model for online multi-object tracking. In: Proceedings of the CVPR, pp 6768–6777

  69. Zhang J, Ma S, Sclaroff S (2014) Meem: robust tracking via multiple experts using entropy minimization. In: Proceedings of the ECCV, pp 188–203. Springer

  70. Zhou B, Khosla A, Lapedriza A, Oliva A, Torralba A (2016) Learning deep features for discriminative localization. In: Proceedings of the CVPR, pp 2921–2929

  71. Zhu J, Yang H, Liu N, Kim M, Zhang W, Yang MH (2018) Online multi-object tracking with dual matching attention networks. In: Proceedings of the ECCV, pp 366–382

  72. Jie Z, Shufang W, Hong Z, Yan L, Li Z (2019) Multi-center convolutional descriptor aggregation for image retrieval. Int J Mach Learn Cybern 10(7):1863–1873

    Article  Google Scholar 

  73. Zhu Z, Wang Q, Li B, Wu W, Yan J, Hu W (2018) Distractor-aware siamese networks for visual object tracking. In: Proceedings of the ECCV, pp 101–117

Download references

Acknowledgements

This work was supported in part by the National Natural Science Foundation of China under Grants 61972097, 61672159, U1705262, and 61672158, in part by the Technology Guidance Project of Fujian Province under Grant 2017H0015, in part by the Natural Science Foundation of Fujian Province under Grants 2021J01612, 2018J1798 and 2018J07005, in part by the Major Project of Fujian Province under Grant 2021HZ022007.

Author information

Author notes

  1. Xiao Ke and Yuezhou Li contributed equally to this work.

Authors and Affiliations

  1. Fujian Provincial Key Laboratory of Networking Computing and Intelligent Information Processing, College of Mathematics and Computer Science, Fuzhou University, Fuzhou, 350116, China

    Xiao Ke, Yuezhou Li, Wenzhong Guo & Yanyan Huang

  2. Key Laboratory of Spatial Data Mining and Information Sharing, Ministry of Education, Fuzhou, 350003, China

    Xiao Ke, Yuezhou Li, Wenzhong Guo & Yanyan Huang

Authors
  1. Xiao Ke
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Yuezhou Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Wenzhong Guo
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Yanyan Huang
    View author publications

    You can also search for this author inPubMed Google Scholar

Corresponding author

Correspondence to Wenzhong Guo.

Ethics declarations

Conflicts of interest

All authors declare that they have no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ke, X., Li, Y., Guo, W. et al. Learning deep convolutional descriptor aggregation for efficient visual tracking. Neural Comput & Applic 34, 3745–3765 (2022). https://doi.org/10.1007/s00521-021-06638-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00521-021-06638-8

Keywords