International journal of advanced smart convergence

The Institute of Internet, Broadcasting and Communication (한국인터넷방송통신학회)
권호 표지
DOI QR코드

DOI QR Code

A Study on the Accuracy Improvement of One-repetition Maximum based on Deep Neural Network for Physical Exercise

  • Lee, Byung-Hoon (Radio & Communications Engineering, Chungbuk National University) ;
  • Kim, Myeong-Jin (Radio & Communications Engineering, Chungbuk National University) ;
  • Kim, Kyung-Seok (Radio & Communications Engineering, Chungbuk National University)
  • Received : 2019.05.04
  • Accepted : 2019.05.17
  • Published : 2019.06.30
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, we conducted a study that utilizes deep learning to calculate appropriate physical exercise information when basic human factors such as sex, age, height, and weight of users come in. To apply deep learning, a method was applied to calculate the amount of fat needed to calculate the amount of one repetition maximum by utilizing the structure of the basic Deep Neural Network. By applying Accuracy improvement methods such as Relu, Weight initialization, and Dropout to existing deep learning structures, we have improved Accuracy to derive a lean body weight that is closer to actual results. In addition, the results were derived by applying a formula for calculating the one repetition maximum load on upper and lower body movements for use in actual physical exercise. If studies continue, such as the way they are applied in this paper, they will be able to suggest effective physical exercise options for different conditions as well as conditions for users.

Keywords

OTNBCL_2019_v8n2_147_f0001.png 이미지

Figure 1. Structure of node

OTNBCL_2019_v8n2_147_f0002.png 이미지

Figure 2. Structure of Deep Neural Network

OTNBCL_2019_v8n2_147_f0003.png 이미지

Figure 3. Result of learning to using sigmoid function

OTNBCL_2019_v8n2_147_f0004.png 이미지

Figure 4. Different of sigmoid and relu function

OTNBCL_2019_v8n2_147_f0005.png 이미지

Figure 5. Result of applying HE initialization to Relu function

OTNBCL_2019_v8n2_147_f0006.png 이미지

Figure 6. Overfitting problem

OTNBCL_2019_v8n2_147_f0007.png 이미지

Figure 8. Result of applying Accuracy improvement techniques

OTNBCL_2019_v8n2_147_f0008.png 이미지

Figure 9. Deep learning output and actual value

OTNBCL_2019_v8n2_147_f0009.png 이미지

Figure 7. (a) Before Dropout, (b) After Dropout

Table 1. 1RM calculation formula by type of exercise using lean body mass

OTNBCL_2019_v8n2_147_t0001.png 이미지

Table 2. Percentage of 1RM according to the purpose of exercise

OTNBCL_2019_v8n2_147_t0002.png 이미지

References

  1. Aarts, Henk, Theo Paulussen, and Herman Schaalma. "Physical exercise habit: on the conceptualization and formation of habitual health behaviours." Health education research 12.3 (1997): 363-374. DOI: https://doi.org/10.1093/her/12.3.363
  2. Ravì, Daniele, et al. "Deep learning for health informatics." IEEE journal of biomedical and health informatics 21.1 (2017): 4-21. https://doi.org/10.1109/JBHI.2016.2636665
  3. Doan, Brandon K., et al. "Effects of increased eccentric loading on bench press 1RM." The Journal of Strength & Conditioning Research 16.1 (2002): 9-13. DOI: https://doi.org/10.1519/1533-4287(2002)016<0009:eoielo>2.0.co;2
  4. Schmidhuber, Jurgen. "Deep learning in neural Networks: An overview." Neural Networks 61 (2015): 85-117. DOI: https://doi.org/10.1016/j.neunet.2014.09.003
  5. Chassiakos, A. G., and S. F. Masri. "Modelling unknown structural systems through the use of neural networks." Earthquake engineering & structural dynamics 25.2 (1996): 117-128. DOI: https://doi.org/10.1002/(sici)1096-9845(199602)25:2<117::aid-eqe541>3.0.co;2-a
  6. Han, Jun, and Claudio Moraga. "The influence of the sigmoid function parameters on the speed of backpropagation learning." International Workshop on Artificial Neural Networks. Springer, Berlin, Heidelberg, 1995. DOI:10.1007/3-540-59497-3_175
  7. Qian, Yao, et al. "On the training aspects of deep neural Network (DNN) for parametric TTS synthesis." 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2014. DOI: https://doi.org/10.1109/icassp.2014.6854318
  8. Kingma, Diederik P., and Jimmy Ba. "Adam: A method for stochastic optimization." arXiv preprint arXiv:1412.6980 (2014).
  9. LeCun, Yann, Yoshua Bengio, and Geoffrey Hinton. "Deep learning." nature 521.7553 (2015): 436. DOI: https://doi.org/10.1038/nature14539
  10. Kumar, Siddharth Krishna. "On weight initialization in deep neural Networks." arXiv preprint arXiv:1704.08863 (2017).
  11. Srivastava, Nitish, et al. "Dropout: a simple way to prevent neural Networks from overfitting." The Journal of Machine Learning Research 15.1 (2014): 1929-1958.
  12. Cadore, E., et al. "Prediction of one repetition maximum load by total and lean body mass in trained and untrained men." Medicina Sportiva 16.3 (2012): 111-117. DOI: https://doi.org/10.5604/17342260.1011391