Skip to main content

Investigation of Methods for Increasing the Efficiency of Convolutional Neural Networks in Identifying Tennis Players

Abstract

The article is devoted to the study of the effectiveness of the convolutional neural networks (CNNs) application for solving the problem of tennis players face recognition. For ease of analysis, two players were selected: Roger Federer (Switzerland) and Rafael Nadal (Spain). To isolate faces from the publicly available images of the players, it is proposed to use the Haar cascades and the Viola–Jones method. These images are used to train and test convolutional networks with various parameters: architecture, including the number of layers; epochs of learning; optimization methods; and also when applying various regularization methods, including drop out and data augmentation. The use of regularization made it possible to reduce the effect of overfitting. In addition, the efficiency of networks with pretrained layers based on transfer learning methods is investigated. The VGG-16 convolutional network is chosen for the transfer learning. For a large number of different combinations of convolutional networks, metrics are calculated for precision, recall, and accuracy. The average gain for these parameters is 25% with the best set of characteristics for convolutional networks and training. It is also shown in the study that the patterns of applying certain modifications are universal for optical images. In particular, similar architectures and training approaches are also tested for the problem of recognizing cats and dogs on a much larger dataset. The study confirms the average increase in recognition metrics of 26%.

This is a preview of subscription content, access via your institution.

Fig. 1.
Fig. 2.
Fig. 3.
Fig. 4.
Fig. 5.
Fig. 6.
Fig. 7.

REFERENCES

  1. N. A. Andriyanov and V. E. Dementiev, “Developing and studying the algorithm for segmentation of simple images using detectors based on doubly stochastic random fields,” Pattern Recognit. Image Anal. 29 (1), 1–9 (2019). https://doi.org/10.1134/S105466181901005X

    Article  Google Scholar 

  2. N. Andriyanov, V. Dementev, A. Tashlinskiy, and K. Vasiliev, “The study of improving the accuracy of convolutional neural networks in face recognition tasks,” Pattern Recognition. ICPR International Workshops and Challenges. ICPR 2021, Ed. by A. Del Bimbo (Springer, Cham, 2021), pp. 5–14. https://doi.org/10.1007/978-3-030-68821-9_1

    Book  Google Scholar 

  3. N. A. Andriyanov, K. K. Vasiliev, and V. E. Dementiev, “Anomalies detection on spatially inhomogeneous polyzonal images,” CEUR Workshop Proc. 1901, 10–15 (2017). https://doi.org/10.18287/1613-0073-2017-1901-10-15

    Article  Google Scholar 

  4. N. A. Andriyanov and D. A. Andriyanov, “The using of data augmentation in machine learning in image processing tasks in the face of data scarcity,” J. Phys.: Conf. Ser. 1661 (1), 012018 (2020).

    Google Scholar 

  5. B. B. Traore, B. Kamsu-Foguem, and F. Tangara, “Deep convolution neural network for image recognition,” Ecol. Inf. 48, 257–268 (2018). https://doi.org/10.1016/j.ecoinf.2018.10.002

    Article  Google Scholar 

  6. A. Buslaev, A. Parinov, E. Khvedchenya, V. Iglovikov, and A. Kalinin, “Albumentations: Fast and flexible image augmentations,” arXiv (2018). arXiv:1809.06839v1 [cs.CV]

  7. M. Coşkun, A. Uçar, Ö. Yıldırım, and Y. Demir, “Face recognition based on convolutional neural network,” in International Conference on Modern Electrical and Energy Systems (IEEE, 2017). https://doi.org/10.1109/MEES.2017.8248937

  8. Guillaume Dave, Xing Chao, and Kishore Sriadibhatla, Face Recognition in Mobile Phones (Stanford Univ., 2010).

    Google Scholar 

  9. V. E. Dementyiev, N. A. Andriyanov, and K. K. Vasilyiev, “Use of images augmentation and implementation of doubly stochastic models for improving accuracy of recognition algorithms based on convolutional neural networks,” in 2020 Systems of Signal Synchronization, Generating and Processing in Telecommunications, SYNCHROINFO 2020 (IEEE, 2020). https://doi.org/10.1109/SYNCHROINFO49631.2020.9166000

    Book  Google Scholar 

  10. Geng Du, Fei Su, and Anni Cai, “Face recognition using SURF features,” in Proceedings of SPIE–The International Society for Optical Engineering (2009). https://doi.org/10.1117/12.832636

  11. https://www.kaggle.com/c/dogs-vs-cats. Accessed March 24, 2021.

  12. M. Fox, “Facial recognition tech secures enterprise access control,” Biom. Technol. Today 2017 (10), 2–3 (2017). https://doi.org/10.1016/S0969-4765(17)30145-5

    Article  Google Scholar 

  13. Ye Li, Yinghui Wang, Jing Liu, and Wen Hao, “Expression-insensitive 3D face recognition by the fusion of multiple subject-specific curves,” Neurocomputing 275, 1295–1307 (2018).

    Article  Google Scholar 

  14. A. J. Logan, G. E. Gordon, and G. Loffler, “Contributions of individual face features to face discrimination,” Vision Res. 137, 29–39 (2017).

    Article  Google Scholar 

  15. Shilpi Singhas and S. V. Prasad, “Techniques and challenges of face recognition: A critical review,” Procedia Comput. Sci. 143, 536–543 (2018). https://doi.org/10.1016/j.procs.2018.10.427

    Article  Google Scholar 

  16. Tanwir Khan, Computer Vision–Detecting Objects Using Haar Cascade Classifier (2019). https://towardsdatascience.com/computer-vision-detecting-objects-using-haar-cascade-classifier-4585472829a9. Accessed March 24, 2021.

  17. K. K. Vasil’ev, V. E. Dement’ev, and N. A. Andriyanov, “Application of mixed models for solving the problem on restoring and estimating image parameters,” Pattern Recognit. Image Anal. 26 (1), 240–247 (2017). https://doi.org/10.1134/S1054661816010284

    Article  Google Scholar 

  18. K. K. Vasiliev and N. A. Andriyanov, “Synthesis and analysis of doubly stochastic models of images,” CEUR Workshop Proc. 2017, 145–154 (2005).

    Google Scholar 

  19. Yang Zhang, Peihua Lv, and Xiaobo Lu, “A deep learning approach for face detection and location on highway,” IOP Conf. Ser.: Mater. Sci. Eng. 435, 012004 (2018). https://doi.org/10.1088/1757-899X/435/1/012004

  20. Zetao Chen, Obadiah Lam, Adam Jacobson, and Michael Milford, “Convolutional neural network-based place recognition,” arXiv (2014). arXiv:1411.1509 [cs.CV]

Download references

Funding

This study was supported by the the Russian Foundation for Basic Research (project no. 19-29-09048) and by the Russian Foundation for Basic Research and the Government of Ulyanovsk oblast (project no. 19-47-730011).

Author information

Authors and Affiliations

Authors

Corresponding authors

Correspondence to N. A. Andriyanov, V. E. Dementev, K. K. Vasiliev or A. G. Tashlinskii.

Ethics declarations

COMPLIANCE WITH ETHICAL STANDARDS

This article is a completely original work by its authors, has not previously been published and will not be published in other publications.

CONFLICT OF INTEREST

The authors declare that they have no conflict of interest.

Additional information

Nikita Andreevich Andriyanov, Candidate of Technical Sciences. Born 1990. Graduated from the Radio Engineering Faculty of Ulyanovsk State Technical University in 2013. Defended his Candidate’s dissertation on the topic “Doubly stochastic autoregressive image models” in 2017. Research interests: pattern recognition, statistical analysis of images, computer vision, and mathematical modeling. Associate Professor at the Department of Data Analysis and Machine Learning at the Financial University under the Government of the Russian Federation.

Vitaliy Evgenevich Dementev, Doctor of Technical Sciences, Associate Professor. Born 1982. Graduated from Ulyanovsk State Technical University with a degree in applied mathematics in 2004. Defended his Candidate’s dissertation on the topic “Signal Detection in Multispectral Images” in 2007. He defended his Doctoral dissertation in 2020. Research interests: statistical analysis of random processes and fields. Head of the Telecommunications Department of Ulyanovsk State Technical University.

Konstantin Konstantinovich Vasiliev, Doctor of Technical Sciences, Professor. Born in 1948. Graduated from Leningrad Electrotechnical Institute with a degree in radioelectronic devices in 1972. Research interests: statistical analysis of random processes and fields. Defended his Candidate’s dissertation in 1975. He defended his Doctoral dissertation in 1985. Honored Worker of Science and Technology of the Russian Federation and Corresponding Member of the Academy of Sciences of the Republic of Tatarstan. Professor of the Department of Telecommunications, Ulyanovsk State Technical University.

Aleksandr Grigor’evich Tashlinskii, Doctor of Technical Sciences, Professor. Born in 1954. Graduated from Ulyanovsk Polytechnic Institute in 1977 with a degree in radio engineering. Research interests: statistical analysis of images, estimation of image parameters, and pseudogradient algorithms. Defended his Doctoral dissertation in 1999. Head of the Department of Radio Engineering, Ulyanovsk State Technical University.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Andriyanov, N.A., Dementev, V.E., Vasiliev, K.K. et al. Investigation of Methods for Increasing the Efficiency of Convolutional Neural Networks in Identifying Tennis Players. Pattern Recognit. Image Anal. 31, 496–505 (2021). https://doi.org/10.1134/S1054661821030032

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1134/S1054661821030032

Keywords:

  • face recognition
  • convolutional neural networks
  • accuracy
  • precision
  • recall
  • probability of correct recognition
  • regularization
  • drop out
  • augmentation
  • doubly stochastic model
  • transfer learning
  • VGG-16
  • Kaggle
  • dogs vs. cats
  • Federer vs. Nadal