Advertisement

System to Capture Movements of Buyers and to Determine Quality of Store Employees

  • A. D. Ulyev
  • V. L. RozalievEmail author
  • Yu. A. Orlova
  • A. V. Alekseev
Conference paper
  • 67 Downloads
Part of the Lecture Notes in Electrical Engineering book series (LNEE, volume 641)

Abstract

The article presents the method of recognition of sales consultants on the basis of a neural network to determine the posture of a person, as well as methods of monitoring the behavior of the sales consultant and analysis of its interaction with the buyer by means of video. The article also discusses the methods of establishing the missing key points on the basis of the physiological structure of the person and the implementation of inter-chamber tracking using key points and segments. A brief overview of similar systems capable of tracking customers or store employees based on video or other related technologies is made. The methods of establishing the identity of the employees of the store and locking the customers to save purchase history. The article presents a description of the proposed methods of operation of the cascade of neural networks to solve the problem, shows the results obtained, as well as the ways of its further improvement.

Keywords

Neural network Artificial intelligence Recognition of human pose Behavior monitoring 

Notes

Acknowledgment

This work was partially supported by RFBR and administration of Volgograd region (grants 17-07-01601, 18-07-00220, 19-47-343001, 19-47-340003, 19-47-340009, 19-07-00020).

References

  1. 1.
    Cao, Z., Simon, T., Wei, S.E. et al.: Real-time multi-person 2D pose estimation using part affinity fields (2018). http://arxiv.org/abs/1611.08050. Accessed 25 July 2019
  2. 2.
    Ulyanova, O.: Psychological features of sales consultants network marketing. SSTU, Samara (2019)Google Scholar
  3. 3.
    Cao, Z.: Real-time multi-person 2D pose estimation using part affinity fields (2018). http://www.ri.cmu.edu/wpcontent/uploads/2017/04/thesis.pdf. Accessed 28 July 2019
  4. 4.
    Iqbal, U., Gall, J.: Multi-person pose estimation with local joint-to-person associations (2018). https://arxiv.org/pdf/1608.08526.pdf. Accessed 25 July 2019
  5. 5.
    Insafutdinov, E., Pishchulin, L., Andres, B., et al.: A deeper, stronger, and faster multi-person pose estimation model (2018). https://arxiv.org/pdf/1605.03170.pdf. Accessed 23 July 2019
  6. 6.
    Osipova, Y., Lavrov, D.: Application of cluster analysis by k-means method for classification of scientific texts (2018). https://cyberleninka.ru/article/n/primenenie-klasternogo-analiza-metodom-k-srednih-dlya-klassifikatsii-tekstov-nauchnoy-napravlennosti. Accessed 29 July 2019
  7. 7.
    Khorunsjiy, M.: Method for quantifying color differences in the perception of digital images. Science, St. Petersburg (2008)Google Scholar
  8. 8.
    Rozaliev, V., Orlova, Y.: Recognition of gesture and poses for the definition of human emotions. In: 11th International Conference of Pattern Recognition and Image Analysis: New Information Technologies Conference proceedings, Samara, 23–28 September 2013, vol. 2, pp. 713–716 (2013)Google Scholar
  9. 9.
    Bobkov, A., Rozaliev, V.: Fuzzification of data describing the movement of a person. open semantic technologies for the design of intelligent systems. In: Materials International Scientific Technology Conference, Minsk, 10–12 February 2011, pp. 483–486 (2011)Google Scholar
  10. 10.
    Black, M., Jacobs, D.: End-to-end recovery of human shape and pose (2018). https://www.researchgate.net/publication/321902575_Endtoend_Recovery_of_Human_Shape_and_Pose?discoverMore=1. Accessed 25 July 2019

Copyright information

© Springer Nature Switzerland AG 2020

Authors and Affiliations

  • A. D. Ulyev
    • 1
  • V. L. Rozaliev
    • 1
    Email author
  • Yu. A. Orlova
    • 1
  • A. V. Alekseev
    • 1
  1. 1.Volgograd State Technical UniversityVolgogradRussian Federation

Personalised recommendations