A Novel KPCA-Based RVM Model for Human Activity Classification
Wu Jianning1*, Lin Qiuting2, Wu Bin3
1(College of Computer and Cyber Security, Fujian Normal University, Fuzhou 350117, China) 2(School of Mathematics and Statistics, Fujian Normal University, Fuzhou 350117, China) 3(Hospital of Fujian Normal University, Fuzhou 350007, China)
Abstract：In order to effectively improve the generalization ability in human activity classification task based on small sample data size, a novel human activity classification model was constructed by using the hybrid technique of kernel principal component (KPCA), and relevance vector machine (RVM) was proposed in this paper. The proposed technique was able to take the advantage of kernel function for the combination of KPCA with RVM, that is, the nonlinear gait features containing more human activity discriminative information hidden in the transformed higher dimensional feature space could be exploited by KPCA, which greatly contributed to solve the sparse probability distribution of human activity discrimination by Bayesian learning algorithm in RVM. This significantly improved the generalization performance of human activity classification. A public UCI human activity recognition (HAR) dataset with wearable sensor data from a smartphone was selected to evaluate the feasibility of our proposed model. In the experiment, all 10 299 of sample data were obtained from the collected data including a total of 30 subjects with six different human activity patterns. The cross-validation with 10 times was used to train and test the model. The experimental results showed that our proposed model could reach the best accuracy of 96% when ten relevance vectors were available, which were 5.4% and 3.6% more than KPCA-SVM and CNN-LSTM deep learning model, respectively. This suggested that our model had a superior ability to extract more nonlinear features associated with human activity change and to learn the sparse probability distribution of RVM. In conclusion, the proposed technique could accurately detect a certain human activity pattern based on the small sample size, which would provide a new idea and approach to exactly identify human activity.
吴建宁, 林秋婷, 伍滨. 基于核主成分分析的相关向量机人体动作分类新型模型[J]. 中国生物医学工程学报, 2022, 41(6): 641-649.
Wu Jianning, Lin Qiuting, Wu Bin. A Novel KPCA-Based RVM Model for Human Activity Classification. Chinese Journal of Biomedical Engineering, 2022, 41(6): 641-649.
[1] Tariq M, Shah MA. Review of model-free gait recognition in biometriesystems[C]//Proceedings of 23rd International Conference on Automation and Computing. Huddersfield: IEEE, 2017: 1-7. [2] JarchiD, Pope J, Lee T, et al. A review on accelerometer-based gait analysis and emerging clinical applications[J]. IEEE Reviews in Biomedical Engineering, 2018, 11:177-194. [3] ManapHH, Tahir NM, Yassin AIM. Statistical analysis of parkinson disease gait classification using Artificial Neural Network[C]// Proceedings of IEEE International Symposium on Signal Processing and Information Technology. Bilbao: IEEE Computer Society, 2011:60-65. [4] Maurer U, Smailagic A, Siewiorek DP, et al. Activity recognition and monitoring using multiple sensors on different body positions[C]//Proceedings of International Workshop on Wearable and Implantable Body Sensor Networks. London: IEEE Computer Society, 2006: 114-116. [5] Bao L, Intille SS. Activity recognition from user-annotated acceleration data[C]//Pervasive Computing. Proceedings, Linz: Springer-Verlag, 2004: 1-17. [6] Brezmes T, Gorricho JL, Cotrina J. Activity recognition from accelerometer data on a mobile phone[C]// Lecture Notes in Computer Science. Salamanca: Springer-Verlag, 2009: 796-799. [7] Manap HH, Tahir NM, Abdullah R. Anomalous gait detection using Naive Bayes classifier[C]// Proceedings of 2012 IEEE Symposium on Industrial Electronics and Applications. Bandung: IEEE, 2012: 378-381. [8] He Zhenyu, JinLianwen. Activity recognition from acceleration data using AR model representation and SVM[C]// Proceedings of the 7thInternational Conference on Machine Learning and Cybernetics. Kunming: IEEE, 2008, 4: 2245-2250. [9] Berchtold M, Budde M, Gordon D, et al.Actiserv: activity recognition service for mobile phones[C]//International Symposium on Wearable Computers 2010.Seoul: IEEE, 2010: 1-8. [10] Chen Changhong, Liang Jimin, Zhu Xiuchnag. Gait recognition based on improved dynamic Bayesian networks[J]. Pattern Recognition, 2011, 44(4): 988-995. [11] Kim YJ, Kang BN, Kim D. Hidden markov model ensemble for activity recognition using tri-axis accelerometer[C]// Proceedings of 2015 IEEE International Conference on Systems, Man, and Cybernetics. Hong Kong: IEEE, 2015: 3036-3041. [12] Hoang T, Choi D, Nguyen T. On the instability of sensor orientation in gait verification on mobile phone[C]//Proceedings of 12th International Joint Conference on e-Business and Telecommunications. Alsace: IEEE, 2015, 4: 148-159. [13] Lemoyne R, Mastroianni T, Cozza M, et al. Implementation of an iPhone as a wireless accelerometer for quantifying gait characteristics[C]//Proceedings of2010 Annual International Conference of the IEEE Engineering in Medicine and Biology. Buenos Aires: IEEE, 2010: 3847-3851. [14] Wang Xiuhui, Zhang Jiajia. Gait feature extraction and gait classification using two-branch CNN[J]. Multimedia Tools and Applications, 2020, 79(3): 2917-2930. [15] Jing Gao, Gu Peishang, Ren Qing, et al. Abnormal gait recognition algorithm based on LSTM-CNN fusion network[J]. IEEE Access, 2019, 7: 163180-163190. [16] Lin Yingjie, WuJianning. A novel multichannel dilated convolution neural network for human activity recognition[J]. Mathematical Problems in Engineering, 2020, 2020:1-11. [17] Avilés C, Ferreyra A, Zúñiga A, et al. Coarse-fine convolutional deep-learning strategy for human activity recognition[J]. Sensors, 2019, 19(1556): 1-16. [18] Tipping ME. The relevance vector machine[C]//Advances in Neural Information Processing Systems.Boston: MIT Press, 2000: 652-658. [19] Zhang Hao, Malik J. Selecting shape features using multi-class relevance vector machine[R]. UCB/EECS-2005-6,2005. [20] Anowar F, Sadaoui S, Selim B. Conceptual and empirical comparison of dimensionality reduction algorithms (PCA, KPCA, LDA, MDS, SVD, LLE, ISOMAP, LE, ICA, t-SNE)[J]. Computer Science Review, 2021, 40(3): 224-229. [21] Wu Jianning, Wang Jue, LiuLi. Feature Extraction via KPCA for classification of gait patterns[J]. Human Movement Science, 2007, 26(3): 393-411. [22] Anguita D, Ghio A, Oneto L, et al. A public domain dataset for human activity recognition using smartphones[C]//Proceedings of 21st European Symposium on Artificial Neural Networks. Bruges: Art Media Press,2013: 437-442. [23] Lavanya B, Gayathri GS. Exploration and deduction of sensor-based humanactivityrecognition system of smart-phone data[C]//Proceedings of 8th IEEE International Conference on Computational Intelligence and Computing Research. Coimbatore: IEEE, 2017: 194-198. [24] Mutegeki R, Han DS. A CNN-LSTM approach to human activity recognition[C]//Proceedings of 2nd International Conference on Artificial Intelligence in Information and Communication.Fukuoka: IEEE, 2020: 362-366. [25] Padmaja BV, Prasa V, Sunitha KVN. A novel random split point procedure using extremely randomized (Extra) trees ensemble method for human activity recognition[J]. EAI Endorsed Transactions on Pervasive Health and Technology, 2020, 6(22): e5.
