生物特徵辨識的使用率近年來廣泛增加，除了指紋、虹膜以及臉部辨識等生物特徵身分辨識方法外，步態身分辨識也是許多研究者投入大量精力研究的一種生物特徵身分辨識方法。每個人的步態數據因為具有難以模仿以及竄改的特性，被視為一種天然的人體加密數據。儘管步態數據擁有許多優勢，進行步態身分辨識時仍然有許多困難與挑戰，例如不同的負重或不同的環境因素等可能會導致個人的步態數據產生變異。本研究使用了四個角落裝有慣性測量單元（IMU）的木質地板進行步態數據收集單步步態，以減少多步步態以及環境因素所可能導致的步態變異。
此研究探討了使用步態數據進行生物特徵識別以及步態階段分類的方法。本研究收集三十位受測者之單步步態數據，其中包括受測者的左右腳三階段步態（起始、連續以及結束），每位受測者在每個狀態下都收集三十步之單步步態。研究中深入比較了不同機器學習模型，如支持向量機（SVM）、隨機森林、XGBoost和深度學習模型（例如卷積神經網絡CNN、長短期記憶模型LSTM以及門控循環單元GRU）在步態身分識別中以及步態階段分類的辨識成效。研究涉及使用裝有慣性測量單元（IMU）的木質地板收集步態數據，並使用各種預處理方法（如移動窗口與連續小波轉換）、各種數據縮放方法（如二元數據轉換、標準化縮放、最大-最小縮放、最大絕對縮放以及魯棒縮放）以及機器學習算法分析這些數據。研究旨在評估不同算法和預處理技術在步態識別和分類中的有效性。
研究結果表明數據預處理方法以及數據縮放對於模型在身分辨識以及步態階段辨識任務上具有顯著影響。在身分辨識任務中，移動窗口搭配無數據縮放或是二元數據縮放的整體表現最佳。移動窗口搭配二元數據縮放之數據使用SVM、1D-CNN、CNN皆有0.99以上的測試準確率，移動窗口搭配無數據縮放在1D-CNN也有0.99以上的測試準確率。在步態階段辨識任務中，無數據預處理搭配無數據縮放在LSTM有最佳準確率，為0.8759。

In recent years, the use of biometric recognition has widely increased. Besides biometric identity recognition methods like fingerprint, iris, and facial recognition, gait identity recognition is also a method many researchers have devoted significant effort to studying. Due to its difficulty in imitating and altering, each person’s gait data is considered a natural form of the human body’s encrypted data. Despite the many advantages of gait data, there are still numerous difficulties and challenges in gait identity recognition, such as variations in individual gait data caused by different loads or environmental factors. This study used a wooden floor equipped with Inertial Measurement Units (IMUs) at four corners for single-step gait data collection, aiming to reduce the gait variations caused by multi-step gaits and environmental factors.
This research explores methods for biometric identification using gait data and for gait phase classification. The study plans to collect single-step gait data from thirty participants, including three phases of gait (initial, continuous, and final) for both left and right feet. Each participant will provide thirty single-step gaits for each phase. The study extensively compares different machine learning models, such as Support Vector Machine (SVM), Random Forest, XGBoost and deep learning models (like Convolutional Neural Networks, Long Short-term Memory and Gated Recurrent Unit) for their effectiveness in gait identity recognition and gait phase classification. The research involves collecting gait data using a wooden floor with IMUs and analyzing these data with various preprocessing methods (like moving windows, binary data transformation, data scaling, etc.) and machine learning algorithms. The study assesses the effectiveness of different algorithms and preprocessing techniques in gait recognition and classification.
Research results indicate that data preprocessing methods and data scaling significantly impact model performance in identity recognition and gait phase recognition tasks. In identity recognition tasks, combining a moving window with no data scaling or binary data scaling performs best overall. The combination of a moving window with binary data scaling achieves a testing accuracy of over 0.99 using SVM, 1D-CNN, and CNN, and the moving window with no data scaling also achieves a testing accuracy of over 0.99 with 1D-CNN. In gait phase recognition tasks, the combination of no data preprocessing with no data scaling achieves the highest accuracy of 0.8759 using LSTM.

Abdulwahab, S. (2017). Deep Learning Models for Paraphrases Identification. MS thesis.
Alzubaidi, L., Zhang, J., Humaidi, A. J., Al-Dujaili, A., Duan, Y., Al-Shamma, O., Santamaría, J., Fadhel, M. A., Al-Amidie, M., & Farhan, L. (2021). Review of deep learning: Concepts, CNN architectures, challenges, applications, future directions. Journal of big Data, 8, 1-74.
Breiman, L. (2001). Random forests. Machine learning, 45, 5-32.
Burges, C. J. (1998). A tutorial on support vector machines for pattern recognition. Data mining and knowledge discovery, 2(2), 121-167.
Chen, J., Wang, Y., Qin, J., Liu, L., & Shao, L. (2017). Fast person re-identification via cross-camera semantic binary transformation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,
Chen, T., & Guestrin, C. (2016). Xgboost: A scalable tree boosting system. Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining,
Chen, Y., Tang, H., Wang, Y., Jin, C., Wang, L., Miao, W., & Wang, X. (2024). The effect of complex cognitive context on the dynamic stability during gait initiation in older women [Original Research]. Frontiers in Aging Neuroscience, 15. https://doi.org/10.3389/fnagi.2023.1342570
Cherdo, Y., Miramond, B., Pegatoquet, A., & Vallauri, A. (2023). Unsupervised Anomaly Detection for Cars CAN Sensors Time Series Using Small Recurrent and Convolutional Neural Networks. Sensors, 23(11), 5013.
Cho, K., Van Merriënboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., & Bengio, Y. (2014). Learning phrase representations using RNN encoder-decoder for statistical machine translation. arXiv preprint arXiv:1406.1078.
Cimolin, V., Cau, N., Galli, M., Pau, M., Parisio, C., Saezza, A., Grugni, G., & Capodaglio, P. (2021). Gait strategy and body composition in patients with Prader–Willi syndrome. Eating and Weight Disorders-Studies on Anorexia, Bulimia and Obesity, 26, 115-124.
Czech, M. D., Psaltos, D., Zhang, H., Adamusiak, T., Calicchio, M., Kelekar, A., Messere, A., Van Dijk, K. R., Ramos, V., & Demanuele, C. (2020). Age and environment-related differences in gait in healthy adults using wearables. NPJ digital medicine, 3(1), 127.
Fazli, S., Askarifar, H., & Tavassoli, M. J. (2011). Gait Recognition using SVM and LDA. Proc. of Int. Conf. on advances in computing, control, and telecommunication technologies,
Figo, D., Diniz, P. C., Ferreira, D. R., & Cardoso, J. M. (2010). Preprocessing techniques for context recognition from accelerometer data. Personal and Ubiquitous Computing, 14, 645-662.
Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural computation, 9(8), 1735-1780.
Jain, A., Hong, L., & Pankanti, S. (2000). Biometric identification. Communications of the ACM, 43(2), 90-98.
Jaskowak, D. J. (2019). Detecting Transient Changes in Gait Using Fractal Scaling of Gait Variability in Conjunction with Gaussian Continuous Wavelet Transform Virginia Tech].
Kim, J., Lee, K. B., & Hong, S. G. (2017). Random forest based-biometric identification using smart shoes. 2017 Eleventh International Conference on Sensing Technology (ICST),
LeCun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11), 2278-2324.
Lesti, G., & Spiegel, S. (2017). A Sliding Window Filter for Time Series Streams. IOTSTREAMING@ PKDD/ECML,
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.
Luo, Z., Chen, J., Cai, X. J., Tanaka, K., Takiguchi, T., Kinkyo, T., & Hamori, S. (2018). Oil price forecasting using supervised GANs with continuous wavelet transform features. 2018 24th International Conference on Pattern Recognition (ICPR),
mehdi Cherrat, E., Alaoui, R., & Bouzahir, H. (2020). Convolutional neural networks approach for multimodal biometric identification system using the fusion of fingerprint, finger-vein and face images. PeerJ Computer Science, 6, e248.
Mileti, I., Germanotta, M., Di Sipio, E., Imbimbo, I., Pacilli, A., Erra, C., Petracca, M., Rossi, S., Del Prete, Z., & Bentivoglio, A. R. (2018). Measuring gait quality in Parkinson’s disease through real-time gait phase recognition. Sensors, 18(3), 919.
Moghaddam, B., & Yang, M.-H. (2002). Learning gender with support faces. IEEE transactions on pattern analysis and machine intelligence, 24(5), 707-711.
Pan, S., Yu, T., Mirshekari, M., Fagert, J., Bonde, A., Mengshoel, O. J., Noh, H. Y., & Zhang, P. (2017). Footprintid: Indoor pedestrian identification through ambient structural vibration sensing. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 1(3), 1-31.
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., & Dubourg, V. (2011). Scikit-learn: Machine learning in Python. the Journal of machine Learning research, 12, 2825-2830.
Qi, J., Ru, F., Ruijie, Z., Jinhua, W., Ting, C., & Fushan, W. (2022). GRU-based multi-scenario gait authentication for smartphones. Chinese Journal of Network & Information Security, 8(5).
Rattanasak, A., Uthansakul, P., Uthansakul, M., Jumphoo, T., Phapatanaburi, K., Sindhupakorn, B., & Rooppakhun, S. (2022). Real-time gait phase detection using wearable sensors for transtibial prosthesis based on a kNN algorithm. Sensors, 22(11), 4242.
Renner, K. E., Peebles, A. T., Socha, J. J., & Queen, R. M. (2022). The impact of sampling frequency on ground reaction force variables. Journal of Biomechanics, 135, 111034.
Research., G. V. (2022). Biometric Technology Market Size Worth $128.09 Billion By 2030 | CAGR: 20.4%. . Grand View Research. https://www.grandviewresearch.com/industry-analysis/biometrics-industry#:~:text=The%20global%20biometric%20technology%20market,in%20consumer%20electronics%20and%20automotive
Sayed, M. (2018). Biometric Gait Recognition Based on Machine Learning Algorithms. J. Comput. Sci., 14(7), 1064-1073.
Semwal, V. B., Mazumdar, A., Jha, A., Gaud, N., & Bijalwan, V. (2021). Speed, cloth and pose invariant gait recognition-based person identification. In Machine learning: theoretical foundations and practical applications (pp. 39-56). Springer.
Shi, B., Yen, S. C., Tay, A., Tan, D. M., Chia, N. S., & Au, W. L. (2020). Convolutional neural network for freezing of gait detection leveraging the continuous wavelet transform on lower extremities wearable sensors data. 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC),
Shi, L.-F., Liu, Z.-Y., Zhou, K.-J., Shi, Y., & Jing, X. (2023). Novel deep learning network for gait recognition using multimodal inertial sensors. Sensors, 23(2), 849.
Shuvo, M. M. H., Ahmed, N., Nouduri, K., & Palaniappan, K. (2020). A hybrid approach for human activity recognition with support vector machine and 1D convolutional neural network. 2020 IEEE Applied Imagery Pattern Recognition Workshop (AIPR),
Świtoński, A., Polański, A., & Wojciechowski, K. (2011). Human identification based on gait paths. Advanced Concepts for Intelligent Vision Systems: 13th International Conference, ACIVS 2011, Ghent, Belgium, August 22-25, 2011. Proceedings 13,
Tanaka, T., Nambu, I., Maruyama, Y., & Wada, Y. (2022). Sliding-window normalization to improve the performance of machine-learning models for real-time motion prediction using electromyography. Sensors, 22(13), 5005.
Vajda, D., Pekar, A., & Farkas, K. (2021). Towards machine learning-based anomaly detection on time-series data. Infocommunications Journal, 13(1), 35-44.
Wu, Z., Huang, Y., Wang, L., Wang, X., & Tan, T. (2016). A comprehensive study on cross-view gait based human identification with deep cnns. IEEE transactions on pattern analysis and machine intelligence, 39(2), 209-226.
Yunas, S. U., & Ozanyan, K. B. (2021). Gait activity classification using multi-modality sensor fusion: a deep learning approach. IEEE Sensors Journal, 21(15), 16870-16879.