Lightweight Action Recognition Free from Detailed Manual Labeling for Edge Computing
DOI:
https://doi.org/10.63282/3050-9262.IJAIDSML-V7I4P103Keywords:
Action Recognition, Wearable Accelerometer, Pseudo-Labeling, Time-Series Clustering, Cross-User Generalization, Lightweight ModelAbstract
This study aims to construct a lightweight action recognition pipeline for work environments that does not rely on detailed manual labeling to achieve both the reduction of manual labeling load in the offline stage and the lightweight recognition of user actions in the online stage. The proposed method generates pseudo-labels from accelerometer data using time-series clustering. Extending the labeling scheme to other subjects, it integrates data from multiple subjects into a common set of act classes. Furthermore, it trains a lightweight action recognition model using the obtained pseudo-labels, allowing the evaluation of the trade-off between model complexity and recognition performance under the assumption of edge deployment. The process verifies not only the transferability of action structures using pseudo-labels, but also the applicability of lightweight models to sequential action recognition.
References
[1] L. Bao and S. S. Intille, "Activity recognition from user-annotated acceleration data," in Pervasive Computing, A. Ferscha and F. Mattern, Eds. Berlin/Heidelberg, Germany: Springer, 2004, vol. 3001, pp. 1-17. Publisher Site
[2] O. Baños, J.-M. Gálvez, M. Damas, H. Pomares, and I. Rojas, "Window size impact in human activity recognition," Sensors, vol. 14, pp. 6474-6499, 2014. Publisher Site
[3] S. Zhang, Y. Li, S. Zhang, F. Shahabi, S. Xia, Y. Deng, and N. Alshurafa, "Deep learning in human activity recognition with wearable sensors: A review on advances," Sensors, vol. 22, Art. no. 1476, 2022. Publisher Site
[4] J. A. Santoyo-Ramón, E. Casilari, and J. M. Cano-García, "A study of the influence of the sensor sampling frequency on the performance of wearable fall detectors," Measurement, vol. 193, Art. no. 110945, 2022. Publisher Site
[5] F. J. Ordóñez and D. Roggen, "Deep convolutional and LSTM recurrent neural networks for multimodal wearable activity recognition," Sensors, vol. 16, Art. no. 115, 2016. Publisher Site
[6] N. Y. Hammerla, S. Halloran, and T. Plötz, "Deep, convolutional, and recurrent models for human activity recognition using wearables," in Proceedings of the Twenty-Fifth International Joint Conference on Artificial Intelligence, New York, NY, USA, Jul. 9-15, 2016, pp. 1533-1540. Publisher Site
[7] H. Zhou, X. Zhang, Y. Feng, T. Zhang, and L. Xiong, "Efficient human activity recognition on edge devices using DeepConv LSTM architectures," Scientific Reports, vol. 15, Art. no. 13830, 2025. Publisher Site
[8] A. O. Ige and M. H. M. Noor, "A survey on unsupervised learning for wearable sensor-based activity recognition," Applied Soft Computing, vol. 127, Art. no. 109363, 2022. Publisher Site
[9] A. Saeed, T. Ozcelebi, and J. Lukkien, "Multi-task self-supervised learning for human activity detection," Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, vol. 3, Art. no. 61, 2019. Publisher Site
[10] A. Logacjov, "Self-supervised learning for accelerometer-based human activity recognition: A survey," Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, vol. 8, Art. no. 149, pp. 1-42, 2024. Publisher Site
[11] H. Haresamudram, I. A. Essa, and T. Plötz, "Contrastive predictive coding for human activity recognition," Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, vol. 5, Art. no. 65, pp. 1-26, 2021. Publisher Site
[12] A. Roy, H. Dutta, A. K. Bhuyan, and S. Biswas, "On-device semi-supervised activity detection: A new privacy-aware personalized health monitoring approach," Sensors, vol. 24, Art. no. 4444, 2024. Publisher Site
[13] H. Yuan, S. Chan, A. P. Creagh, et al., "Self-supervised learning for human activity recognition using 700,000 person-days of wearable data," npj Digital Medicine, vol. 7, Art. no. 91, 2024. Publisher Site
[14] C. I. Tang, I. Pérez-Pozuelo, D. Spathis, S. Brage, N. Wareham, and C. Mascolo, "SelfHAR: Improving human activity recognition through self-training with unlabeled data," Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, vol. 5, Art. no. 36, 2021. Publisher Site
[15] R. Fallahzadeh and H. Ghasemzadeh, "Personalization without user interruption: Boosting activity recognition in new subjects using unlabeled data," in Proceedings of the 2017 ACM/IEEE 8th International Conference on Cyber-Physical Systems, Pittsburgh, PA, USA, Apr. 18-21, 2017, pp. 293-302. Publisher Site
[16] O. Baños, M. A. Toth, M. Damas, H. Pomares, and I. Rojas, "Dealing with the effects of sensor displacement in wearable activity recognition," Sensors, vol. 14, pp. 9995-10023, 2014. Publisher Site
[17] A. Rosales Sanabria, F. Zambonelli, S. Dobson, and J. Ye, "ContrasGAN: Unsupervised domain adaptation in human activity recognition via adversarial and contrastive learning," Pervasive and Mobile Computing, vol. 78, Art. no. 101477, 2021. Publisher Site
[18] Z. Fu, X. He, E. Wang, J. Huo, J. Huang, and D. Wu, "Personalized human activity recognition based on integrated wearable sensor and transfer learning," Sensors, vol. 21, Art. no. 885, 2021. Publisher Site
[19] R. Hu, L. Chen, S. Miao, and X. Tang, "SWL-Adapt: An unsupervised domain adaptation model with sample weight learning for cross-user wearable human activity recognition," Proceedings of the AAAI Conference on Artificial Intelligence, vol. 37, pp. 6012-6020, 2023. Publisher Site
[20] J. MacQueen, "Some methods for classification and analysis of multivariate observations," in Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, Berkeley, CA, USA, 1967, vol. 1, pp. 281-297. Publisher Site
[21] A. P. Dempster, N. M. Laird, and D. B. Rubin, "Maximum likelihood from incomplete data via the EM algorithm," Journal of the Royal Statistical Society: Series B, vol. 39, pp. 1-38, 1977. Publisher Site
[22] E. Keogh and J. Lin, "Clustering of time-series subsequences is meaningless: Implications for previous and future research," Knowledge and Information Systems, vol. 8, pp. 154-177, 2005. Publisher Site
[23] K. Li, R. Habre, H. Deng, R. Urman, J. Morrison, F. D. Gilliland, et al., "Applying multivariate segmentation methods to human activity recognition from wearable sensors' data," JMIR mHealth and uHealth, vol. 7, Art. no. e11201, 2019. Publisher Site
[24] D. Guijo-Rubio, A. M. Durán-Rosal, P. A. Gutiérrez, A. Troncoso, and C. Hervás-Martínez, "Time-series clustering based on the characterization of segment typologies," IEEE Transactions on Cybernetics, vol. 51, pp. 5409-5422, 2021. Publisher Site
[25] D. Hallac, S. Vare, S. Boyd, and J. Leskovec, "Toeplitz inverse covariance-based clustering of multivariate time series data," in Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Halifax, NS, Canada, Aug. 13-17, 2017, pp. 215-223. Publisher Site
[26] P. Agarwal and M. Alam, "A lightweight deep learning model for human activity recognition on edge devices," Procedia Computer Science, vol. 167, pp. 2364-2373, 2020. Publisher Site
[27] V. Sharma, D. Pau, and J. Cano, "Efficient tiny machine learning for human activity recognition on low-power edge devices," in Proceedings of the 2024 IEEE 8th International Forum on Research and Technologies for Society and Industry (RTSI), Milano, Italy, Sep. 18-20, 2024, pp. 85-90. Publisher Site
[28] L. Breiman, "Random forests," Machine Learning, vol. 45, pp. 5-32, 2001. Publisher Site
[29] F. Daghero, A. Burrello, C. Xie, L. Benini, A. Calimera, E. Macii, M. Poncino, and D. Jahier Pagliari, "Adaptive random forests for energy-efficient inference on microcontrollers," in Proceedings of the 2021 IFIP/IEEE 29th International Conference on Very Large Scale Integration, Singapore, Oct. 4-8, 2021, pp. 1-6. Publisher Site
[30] F. Küppers, J. Albers, and A. Haselhoff, "Random forest on an embedded device for real-time machine state classification," in Proceedings of the 2019 27th European Signal Processing Conference, A Coruña, Spain, Sep. 2-6, 2019, pp. 1-5. Publisher Site
[31] T. Chen and C. Guestrin, "XGBoost: A scalable tree boosting system," in Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA, Aug. 13-17, 2016, pp. 785-794. Publisher Site
[32] K. H. Brodersen, C. S. Ong, K. E. Stephan, and J. M. Buhmann, "The balanced accuracy and its posterior distribution," in Proceedings of the 20th International Conference on Pattern Recognition, Istanbul, Turkey, Aug. 23-26, 2010, pp. 3121-3124. Publisher Site
