Ensemble Machine Learning for Marker-Free GAIT-Based Human Authentication

Authors

Amogha B
REVA University image/svg+xml
Rohini Deshpande
Reva University, Rukmini Knowledge park, Yelahanka, Bengaluru- 560064, Karnataka, India
Prameela Kumari N
REVA University image/svg+xml
DOI: https://doi.org/10.52549/ijeei.v14i1.7044
Download PDF
Published: 26-03-2026 • Issue: Vol. 14 No. 1: March 2026 • Section: Regular paper

Abstract

The pattern of walking is called as GAIT. This pattern can be used to authenticate persons from a distance. I-GAIT is a method where Indian people are identified and authenticated using a walking pattern. Biometric-based authentication systems are designed to authenticate people using their behavioural and physical patterns. After the COVID-19 pandemic, people are adopting to contactless authentication systems. GAIT is in great demand as authentication is contactless. This paper proposes a methodology to authenticate 240 persons who walked indoors and outdoors in a complex environment. Each person is made to walk in three different situations, such as normal walking (NW), holding a bag (HG), and wearing a coat (WC). Authentication is achieved by calculating the distance between two, four, five and cross-body human joints and morphological features. The distance of the human body is determined by extracting the landmarks from the colour image of a person using mediapipe. Features are trained using five machine learning methods such as random forest, XGboost, LightGBM, gradient boosting and logistic regression. Human recognition is performed by using hard voting, where the majority voted human class is provided as the final predicted class. Authentication accuracy to indoor database is 83%, and for the outdoor database is 86%.
Keywords: I-GAIT; Authentication; Mediapipe; Machine learning; Morphology

References

Rani, V., & Kumar, M. (2023). Human GAIT recognition: A systematic review. Multimedia Tools and Applications, 82(24), 37003-37037. DOI: https://doi.org/10.1007/s11042-023-15079-5

Khosravy, M., Nakamura, K., Hirose, Y., Nitta, N., & Babaguchi, N. (2022). Model inversion attack by integration of deep generative models: Privacy-sensitive face generation from a face recognition system. IEEE Transactions on Information Forensics and Security, 17, 357-372. DOI: https://doi.org/10.1109/TIFS.2022.3140687

Li, H., Qiu, Y., Zhao, H., Zhan, J., Chen, R., Wei, T., & Huang, Z. (2022). GAITSlice: A GAIT recognition model based on spatio-temporal slice features. Pattern Recognition, 124, 108453. DOI: https://doi.org/10.1016/j.patcog.2021.108453

Liao, R., Yu, S., An, W., & Huang, Y. (2020). A model-based GAIT recognition method with body pose and human prior knowledge. Pattern Recognition, 98, 107069. DOI: https://doi.org/10.1016/j.patcog.2019.107069

Gupta, S. K., & Chattopadhyay, P. (2021). GAIT recognition in the presence of co-variate conditions. Neurocomputing, 454, 76-87. DOI: https://doi.org/10.1016/j.neucom.2021.04.113

Faizan, Saleem., Muhammad, Attique, Khan., Majed, Alhaisoni., Usman, Tariq., Ammar, Armghan., Fayadh, Alenezi., Jung-In, Choi., Seifedine, Kadry. (2021). Human GAIT Recognition: A Single Stream Optimal Deep Learning Features Fusion. Sensors, DOI: https://doi.org/10.3390/s21227584

Byungjoo, Noh., Changhong, Youm., Eunkyoung, Goh., Myeounggon, Lee., Hwayoung, Park., Hyojeong, Jeon., Oh, Yoen, Kim. (2021). XGBoost based machine learning approach to predict the risk of fall in older adults using GAIT outcomes.Scientific Reports, DOI: https://doi.org/10.1038/s41598-021-91797-w

Balazia, M., & Sojka, P. (2016, December). Learning robust features for GAIT recognition by maximum margin criterion. In 2016 23rd International Conference on Pattern Recognition (ICPR) (pp. 901-906).IEEE. DOI: http://dx.doi.org/10.1109/ICPR.2016.7899750

Sharif, M. I., Khan, M. A., Alqahtani, A., Nazir, M., Alsubai, S., Binbusayyis, A., & DamaÅ¡eviÄius, R. (2022). Deep learning and kurtosis-controlled, entropy-based framework for human GAIT recognition using video sequences. Electronics, 11(3), 334. DOI: https://doi.org/10.3390/electronics11030334

Nemes, S., & Antal, M. (2021, May). Feature learning for accelerometer based GAIT recognition. In 2021 IEEE 15th International Symposium on Applied Computational Intelligence and Informatics (SACI) (pp. 479-484). IEEE. DOI: https://doi.org/10.1109/SACI51354.2021.9465576

Maria, De, Marsico., Eduard, Gabriel, Fartade., Alessio, Mecca. (2018). Feature-based Analysis of GAIT Signals for Biometric Recognition - Automatic Extraction and Selection of Features from Accelerometer Signals. 630-637. DOI: https://doi.org/10.5220/0006719106300637

Sabir, A., Al-jawad, N., & Jassim, S. (2013, May). GAIT recognition using spatio-temporal silhouette-based features. In Mobile Multimedia/Image Processing, Security, and Applications 2013 (Vol. 8755, pp. 194-203). SPIE. DOI: https://ui.adsabs.harvard.edu/link_gateway/2013SPIE.8755E..0RS/doi:10.1117/12.2017950

Mogan, J. N., Lee, C. P., & Lim, K. M. (2020, March). GAIT recognition using histograms of temporal gradients. In Journal of Physics: Conference Series (Vol. 1502, No. 1, p. 012051). IOP Publishing. DOI: 10.1088/1742-6596/1502/1/012051

S. Sivakumar, A. A. Gopalai, D. Gouwanda and K. H. Lim, "Estimation Of Joint Angle From Ground Reaction Force In Human Gait," 2018 IEEE-EMBS Conference on Biomedical Engineering and Sciences (IECBES), Sarawak, Malaysia, 2018, pp. 623-628, DOI: 10.1109/IECBES.2018.8626650.

Yang, Chaoyu, Renjie Geng, and Qian Liu. "Gait Feature Recognition Based on Multi-instance Learning." (2015). DOI: https://pos.sissa.it/259/084/pdf

Martins, I., Carvalho, P., Corte-Real, L., & Alba-Castro, J. L. (2018). Bmog: boosted gaussian mixture model with controlled complexity for background subtraction. Pattern Analysis and Applications, 21, 641-654. DOI: https://link.springer.com/article/10.1007/s10044-018-0699-y

Kondragunta, J., & Hirtz, G. (2020, July). GAIT parameter estimation of elderly people using 3D human pose estimation in early detection of dementia. In 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC) (pp. 5798-5801). IEEE. DOI: https://doi.org/10.1109/embc44109.2020.9175766

Chen, K. Y., Shin, J., Hasan, M. A. M., Liaw, J. J., Yuichi, O., & Tomioka, Y. (2022). Fitness movement types and completeness detection using a transfer-learning-based deep neural network. Sensors, 22(15), 5700. DOI: https://doi.org/10.3390/s22155700

Lafortune, M. A., Cavanagh, P. R., Sommer Iii, H. J., & Kalenak, A. (1992). Three-dimensional kinematics of the human knee during walking. Journal of biomechanics, 25(4), 347-357. DOI: https://doi.org/10.1016/0021-9290(92)90254-x

Lang, S., Murrow, G., Lang, S., & Murrow, G. (1988). The Distance Formula. Geometry: A High School Course, 110-122.

Yousef, R. N., Khalil, A. T., Samra, A. S., & Ata, M. M. (2023). Model-based and model-free deep features fusion for high performed human GAIT recognition. The Journal of Supercomputing, 79(12), 12815-12852. DOI: https://doi.org/10.1007/s11227-023-05156-9

“Elbow Imageâ€, available at, https://www.vrogue.co/post/elbow-anatomy-upper-limb-anatomy-anatomy-bones-anatomy-study

Comer, M. L., & Delp III, E. J. (1999). Morphological operations for color image processing. Journal of electronic imaging, 8(3), 279-289. DOI: https://doi.org/10.1117/1.482677

Tasel, S. F., Mumcuoglu, E. U., Hassanpour, R. Z., & Perkins, G. (2016). A validated active contour method driven by parabolic arc model for detection and segmentation of mitochondria. Journal of structural biology, 194(3), 253-271. DOI: https://doi.org/10.1016/j.jsb.2016.03.002

“Major axis and minor axisâ€, available at, https://courses.lumenlearning.com/suny-osalgebratrig/chapter/the-ellipse/

Yang, Y., & Liu, X. (1999, August). A re-examination of text categorization methods. In Proceedings of the 22nd annual international ACM SIGIR conference on Research and development in information retrieval (pp. 42-49). DOI: https://doi.org/10.1145/312624.312647

Author Biographies

Amogha B
I am research scholar of Reva university in Electronics and communication department. My personal keywords are, image processsing, feature optimization, computer vision, pattern recognitation, machine and deep learning. Currently working on authentication of persons using walking style.
Rohini Deshpande
Dr. Rohini Deshpande received her Ph. D degree in Electronics and Communication Engg., from Guru Gobind Singh Indraprasth University (GGSIPU) New Delhi, M E degree in Electronics and Communication Engg, from the University of Delhi, with honors and Bachelor degree in Electronics and Communication Engg, from Gulbarga Univeristy. She is in academics for the last 32 years and involved in research and teaching UG and PG students. Her major areas of research are Signal Processing, Optical Networks, Mobile Computation etc.She has published more than 36 papers in journals and conferences of good impact factors. She has citation index of 69 and h-index of 5. She has given invited talk in various occasions including many international conferences. She also has administrative experience of around 10 years. She has delivered the responsibilities of head of ECE, IT, CSE and Mechatronics Engineering departments. She has research experience of 17 years. She has guided one scholar and is guiding five scholars for their doctoral degrees. Presently she is professor in school of ECE, REVA University, Bengaluru.
Prameela Kumari N
Received PhD in Electronics and communication Engineering. Specialization in Nanoelectronics. Area of interest is in VLSI Nanoelectronics Robotics. Having 12 years of teaching experience. Posses two patents

How to Cite

[1]
A. B, R. Deshpande, and P. Kumari N, “Ensemble Machine Learning for Marker-Free GAIT-Based Human Authentication”, IJEEI, vol. 14, no. 1, pp. 196–210, Mar. 2026, doi: 10.52549/ijeei.v14i1.7044.