MLP Tabanlı DNN Modeli Kullanılarak Akıllı Alanlar İçin Yürüyüş Analizinden Kişi Tanıma

Abstract

Akıllı alanlarda, teknoloji kullanılarak oluşabilecek tehditlere karşı insanları korumak ve kriz yönetimi sağlamak için güvenlik önlemleri alınmakta, alan güvenliğinin ölçülmesi ve etkinliğinin sağlanması işlevleri yerine getirilmektedir. Bu ölçümün bir unsuru olarak gelecekte kişi tanımanın en önemli faktör olabileceği düşünülmektedir. Birçok veri ile hızlı ve yüksek doğrulukta sonuçlar verebilen derin öğrenme tabanlı algoritmaların bugün olduğu gibi gelecekte de bu sektörün ayrılmaz bir parçası olacağı görülmektedir. Ancak literatür incelendiğinde bu yöndeki çalışmaların başarısını artırmak için Derin öğrenme algoritmalarının kullanıldığı araştırma sayısının ve sistem pratikliğinin yetersiz olduğu anlaşılmaktadır. Bu nedenle bu çalışmada giyilebilir sensörler sayesinde elde edilen 15 kişinin yürüme verileri kullanılarak insanları tanımak için derin öğrenme kullanılmıştır. Veri çeşitliliğindeki artış, oluşturulan modelin öğrenilmesini olumlu etkileyeceğinden veri artırması yapılmış ve bu veriler MLP tabanlı DNN modelinde sınıflandırılmıştır. Sonuçlar istatistiksel olarak analiz edilmesi sonucunda önerilen modelin yürüme verilerinden kişi tanımada mükemmel performans sergilediğini göstermiştir. Ayrıca ACC oranı %100 bulunmuş ve verileri artırmak için kullanılan yöntemin yürüme verilerinde de başarılı sonuçlar ürettiği kanıtlanmıştır. Çalışmanın başarısının literatürde akıllı alanlara yönelik yeni çalışmalara önemli bir perspektif desteği sağlayabileceği düşünülmektedir.

Keywords

• Yapay zeka
• Derin sinir ağları
• Yürüyüş analizi
• Kişi tanıma
• Akıllı alanlar

Person Recognition from Gait Analysis for Smart Spaces by using MLP-based DNN model

Abstract

In smart fields, security measures are taken to protect people against threats that may arise by using technology and to provide crisis management, and the functions of measuring area security and ensuring its effectiveness are carried out. As an element of this measurement, it is thought that person recognition may be the most important factor in the future. It is seen that deep learning-based algorithms, which can provide fast and high-accuracy results with many data, will be an integral part of this sector in the future as they are today. However, when the literature is examined, it is understood that the number of research in which Deep learning algorithms are used in order to increase the success of the studies in this direction and the system practicality is insufficient. For this reason, in this study, deep learning was used to recognize people by using the walking data of 15 people obtained thanks to wearable sensors. Since the increase in the diversity of the data will positively affect the learning of the created model, data augmentation has been made and these data have been classified in the MLP-based DNN model. The results were statistically analyzed and showed that this model exhibited excellent performance in person recognition from walking data. In addition, the ACC rate was found to be 100%, and it proved that the method used to increase the data also produced successful results in walking data. It is thought that the success of the study can provide important perspective support to new studies for smart fields in the literature.

Keywords

• Artificial Intelligence
• Deep Neural Networks
• Gait Analysis
• Person Recognition
• Smart Fields

References

1. [1] D. Rothman, Artificial Intelligence by Example: Develop machine intelligence from scratch using real artificial intelligence use cases. Packt Publishing Ltd, 2018.
2. [2] M. Molinara, A. Bria, S. De Vito, and C. Marrocco, "Artificial intelligence for distributed smart systems," vol. 142, ed: Elsevier, 2021, pp. 48-50.
3. [3] C. Su, Z. Xu, J. Pathak, and F. Wang, "Deep learning in mental health outcome research: a scoping review," Translational Psychiatry, vol. 10, no. 1, pp. 1-26, 2020.
4. [4] M. A. Al-Garadi, A. Mohamed, A. K. Al-Ali, X. Du, I. Ali, and M. Guizani, "A survey of machine and deep learning methods for internet of things (IoT) security," IEEE Communications Surveys & Tutorials, vol. 22, no. 3, pp. 1646-1685, 2020.
5. [5] P. Ping, Y. Sheng, W. Qin, C. Miyajima, and K. Takeda, "Modeling driver risk perception on city roads using deep learning," IEEE Access, vol. 6, pp. 68850-68866, 2018.
6. [6] A. Makkar and N. Kumar, "An efficient deep learning-based scheme for web spam detection in IoT environment," Future Generation Computer Systems, vol. 108, pp. 467-487, 2020.
7. [7] X. Sun, K. Su, and C. Fan, "VFL—A deep learning-based framework for classifying walking gaits into emotions," Neurocomputing, vol. 473, pp. 1-13, 2022.
8. [8] F. Duan, Y. Lv, Z. Sun, and J. Li, "Multi-Scale Learning for Multimodal Neurophysiological Signals: Gait Pattern Classification as An Example," Neural Processing Letters, 2022.

9. [9] S. Krutaraniyom, K. Sengchuai, A. Booranawong, and J. Jaruenpunyasak, "Pilot Study on Gait Classification Using Machine Learning," in 2022 International Electrical Engineering Congress (iEECON), 2022, pp. 1-4: IEEE.
10. [10] M. Lee, J.-H. Lee, and D.-H. Kim, "Gender recognition using optimal gait feature based on recursive feature elimination in normal walking," Expert Systems with Applications, vol. 189, p. 116040, 2022.
11. [11] K. Delac and M. Grgic, "A survey of biometric recognition methods," in Proceedings. Elmar-2004. 46th International Symposium on Electronics in Marine, 2004, pp. 184-193: IEEE.
12. [12] J. Nelson, "Access control, access badges, and biometrics characteristics for schools," in The Handbook for School Safety and Security: Elsevier, 2014, pp. 241-253.
13. [13] A. Gumuscu, "Improvement of wearable gait analysis sensor based human classification using feature selection algorithms," Fırat Üniversitesi Mühendislik Bilimleri Dergisi, vol. 31, pp. 463-471, 2019.
14. [14] A. Gümüşçü, "Wearable Sensor based Gait Recognition for Human Identification," IMESET’18 DUBAI, p. 31, 2018.
15. [15] X. Li, H. Xu, and J. T. Cheung, "Gait-force model and inertial measurement unit-based measurements: A new approach for gait analysis and balance monitoring," Journal of Exercise Science & Fitness, vol. 14, no. 2, pp. 60-66, 2016.
16. [16] A. Saboor et al., "Latest research trends in gait analysis using wearable sensors and machine learning: A systematic review," Ieee Access, vol. 8, pp. 167830-167864, 2020.
17. [17] D. Dua and C. Graff, "UCI Machine Learning Repository [http://archive.ics.uci.edu/ml]. Irvine, CA: University of California," School of Information and Computer Science, vol. 25, p. 27, 2019.
18. [18] S. E. Dreyfus, "Artificial neural networks, back propagation, and the Kelley-Bryson gradient procedure," Journal of guidance, control, and dynamics, vol. 13, no. 5, pp. 926-928, 1990.
19. [19] J. L. Balcázar, R. Gavalda, and H. T. Siegelmann, "Computational power of neural networks: A characterization in terms of Kolmogorov complexity," IEEE Transactions on Information Theory, vol. 43, no. 4, pp. 1175-1183, 1997.
20. [20] C. M. Bishop, Neural networks for pattern recognition. Oxford university press, 1995.
21. [21] H. Taud and J. Mas, "Multilayer perceptron (MLP)," in Geomatic approaches for modeling land change scenarios: Springer, 2018, pp. 451-455.
22. [22] Y. Bengio, Learning deep architectures for AI. Now Publishers Inc, 2009.
23. [23] J. Schmidhuber, "Deep learning in neural networks: An overview," Neural networks, vol. 61, pp. 85-117, 2015.
24. [24] C. Szegedy, A. Toshev, and D. Erhan, "Deep neural networks for object detection," Advances in neural information processing systems, vol. 26, 2013.
25. [25] R. D. Hof, "Is artificial intelligence finally coming into its own," ed: MIT Technology Review. Accessed, 2018.
26. [26] G. E. Hinton, "A practical guide to training restricted Boltzmann machines," in Neural networks: Tricks of the trade: Springer, 2012, pp. 599-619.
27. [27] Y. You, A. Buluç, and J. Demmel, "Scaling deep learning on gpu and knights landing clusters," in Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, 2017, pp. 1-12.
28. [28] A. Viebke, S. Memeti, S. Pllana, and A. Abraham, "CHAOS: a parallelization scheme for training convolutional neural networks on Intel Xeon Phi," The Journal of Supercomputing, vol. 75, no. 1, pp. 197-227, 2019.
29. [29] D. P. Kingma and J. Ba, "Adam: A method for stochastic optimization," arXiv preprint arXiv:1412.6980, 2014.
30. [30] Y. Ma, L. Guo, and B. Cukic, "A statistical framework for the prediction of fault-proneness," in Advances in Machine Learning Applications in Software Engineering: IGI Global, 2007, pp. 237-263.
31. [31] D. M. Powers, "Evaluation: from precision, recall and F-measure to ROC, informedness, markedness and correlation," arXiv preprint arXiv:2010.16061, 2020.
32. [32] Ş. Yücelbaş, C. Yücelbaş, G. Tezel, S. Özşen, and Ş. Yosunkaya, "Automatic sleep staging based on SVD, VMD, HHT and morphological features of single-lead ECG signal," Expert Systems with Applications, vol. 102, pp. 193-206, 2018.
33. [33] Ş. Yücelbaş, C. Yücelbaş, G. Tezel, S. Özşen, S. Küççüktürk, and Ş. Yosunkaya, "Pre-determination of OSA degree using morphological features of the ECG signal," Expert Systems with Applications, vol. 81, pp. 79-87, 2017.