Analysis of ESP32 SoC for Feed-Forward Neural Network Applications

  • Kristian DokicEmail author
  • Dubravka Mandusic
  • Bojan Radisic
Conference paper
Part of the Learning and Analytics in Intelligent Systems book series (LAIS, volume 7)


In the last few years, microcontrollers used for IoT devices became more and more powerful, and many authors have started to use them in machine learning systems. Most of the authors used them just for data collecting for ML algorithms in the clouds, but some of them implemented ML algorithms on the microcontrollers. The goal of this paper is to analyze the speed of one popular SoC (Espressif ESP32) running a neural network application. Neural networks with one and two hidden layers are used with the different number of neurons and the different number of inputs (9, 36, 144 and 576). This SoC has been analysed because some companies started to produce them with UXGA camera implemented and it can be used to distribute computing load from central ML servers to them. They can be so-called “edge” devices and in this paper speed of ESP32 SoC in feed-forward mode has been analysed.


ESP32 ESP32-CAM Neural network Machine learning 


  1. 1.
    Li, H., Ota, K., Dong, M.: Learning IoT in edge: deep learning for the Internet of Things with edge computing, IEEE Network, pp. 96–101 (2018)CrossRefGoogle Scholar
  2. 2.
    Espressif System, ESP32 Overview. Accessed 15 June 2019Google Scholar
  3. 3.
    Biswas, S.B., Iqbal, M.T.: Solar water pumping system control using a low cost ESP32 microcontroller. In: IEEE Canadian Conference on Electrical & Computer Engineering, Quebec City (2018)Google Scholar
  4. 4.
    Abdullah, A.H., et al.: Development of ESP32-based Wi-Fi electronic nose system for monitoring LPG leakage at gas cylinder refurbish plant. In: 2018 International Conference on Computational Approach in Smart Systems Design and Applications (ICASSDA), Kuching (2018)Google Scholar
  5. 5.
    Maier, A., Sharp, A., Vagapov, Y.: Comparative analysis and practical implementation of the ESP32 microcontroller module for the Internet of Things. In: 7th IEEE International Conference on Internet Technologies, Wrexham (2017)Google Scholar
  6. 6.
    Espressif System, ESP-WHO. Accessed 19 June 2019
  7. 7.
    Allan, A.: Face detection and recognition on the ESP32, 10 December 2018. Accessed 7 June 2019
  8. 8.
    Chand, G., Ali, M., Barmada, B., Liesaputra, V., Ramirez-Prado, G.: Tracking a Person’s Behaviour in a Smart House. In: ICSOC 2018: Service-Oriented Computing – ICSOC 2018 Workshops, Hangzhou (2019)Google Scholar
  9. 9.
    Fernoaga, V.P., Stelea, G.-A., Balan, A., Sandu, F.: OCR-based solution for the integration of legacy and-or non-electric counters in cloud smart grids. In: IEEE 24th International Symposium for Design and Technology in Electronic Packaging (SIITME), Iași (2018)Google Scholar
  10. 10.
    Islam Chowdhuryy, M.H., Sultana, M., Ghosh, R., Ahamed, J.U., Mahmood, M.: AI assisted portable ECG for fast and patient specific diagnosis. In: International Conference on Computer, Communication, Chemical, Materials and Electronic Engineering, Rajshahi (2018)Google Scholar
  11. 11.
    Rosato, D., Masciadri, A.: Non-invasive monitoring system to detect siting people. In: Goodtechs, Bologna (2018)Google Scholar
  12. 12.
    Komarek, A., Pavlik, J., Mercl, L., Sobeslav, V.: Hardware layer of ambient intelligence environment implementation. In: Conference on Computational Collective Intelligence Technologies and Applications (2017)CrossRefGoogle Scholar
  13. 13.
    Zidek, K., Janacova, D., Hosovsky, A., Pitel, J., Lazorik, P.: Data optimization for communication between wireless IoT devices and cloud platforms in production process. In: 3rd EAI International Conference on Management of Manufacturing Systems, Dubrovnik (2018)Google Scholar
  14. 14.
    Kokoulin, A.N., Tur, A.I., Yuzhakov, A.A., Knyazev, A.I.: Hierarchical convolutional neural network architecture in distributed facial recognition system. In: IEEE Conference of Russian Young Researchers in Electrical and Electronic Engineering (EIConRus), Saint Petersburg and Moscow (2019)Google Scholar
  15. 15.
    Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., Chen, L.-C.: MobileNetV2: inverted residuals and linear bottlenecks. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City (2018)Google Scholar
  16. 16.
    Zhang, K., Zhang, Z., Li, Z., Qiao, Y.: Joint face detection and alignment using multi-task cascaded convolutional networks. IEEE Signal Process. Lett. 23(10), 1499–1503 (2016)CrossRefGoogle Scholar
  17. 17.
    Ortega-Zamorano, F., Subirats, J.L., Molina, I., Franco, L., Jerez, J.M.: Smart sensor/actuator node reprogramming in changing environments using a neural network model. Engineering Applications of Artificial Intelligence (2014)Google Scholar
  18. 18.
    Rai, N., Rai, B.: Neural network based closed loop speed control of DC motor using Arduino Uno. Int. J. Eng. Trends Technol. 4(2), 137–140 (2013)Google Scholar
  19. 19.
    Adhitya, R., et al.: Comparison methods of fuzzy logic control and feed forward neural network in automatic operating temperature and humidity control system (Oyster Mushroom Farm House) using microcontroller. In: International Symposium on Electronics and Smart Devices (ISESD), Bandung (2016)Google Scholar
  20. 20.
    RussBarr, Unexpectedly low floating-point performance in C, 13 January 2017. Accessed 22 September 2019
  21. 21.
    Ali, A., Zanzinger, Z., Debose, D., Stephens, B.: Open source building science sensors (OSBSS): a low-cost Arduino-based platform for long-term indoor environmental data collection. Build. Environ. 100, 114–126 (2016)CrossRefGoogle Scholar
  22. 22.
    LeCun, Y.: THE MNIST DATABASE of handwritten digits, Courant Institute. Accessed 12 June 2019

Copyright information

© Springer Nature Switzerland AG 2020

Authors and Affiliations

  1. 1.Polytechnic in PozegaPozegaCroatia
  2. 2.Faculty of AgricultureZagrebCroatia

Personalised recommendations