Abstract
We propose a model that combines only simple techniques to meet the challenge of cooking activity recognition. The challenge dataset is basically small, consisting only of four subjects where three are used for training and one for validation. In order not to overfit the small training data, we employed two simple classifiers, LightGBM and Naive Bayes, which suited the task. To prevent leakage from other subject data during training, we used Leave One Subject Out cross validation. Further, we incorporated a post-processing step wherein the Naive Bayes corrects the macro-activity classification outcomes that have been derived by LightGBM, based on the combinations of macro and micro activities that are likely to occur. We hypothesized that this added post-processing will improve the macro-activity recognition, and with it, our model may be able to adapt well and generalize to other small datasets. As a result, our proposed model achieved an average accuracy of 0.557 when classifying macro and micro activities from a small dataset.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Vrigkas, M., Nikou, C., Kakadiaris, I.A.: A review of human activity recognition methods. Front. Robot. AI 2, 28 (2015). https://doi.org/10.3389/frobt.2015.00028.
Zhang, S., Wei, Z., Nie, J., Huang, L., Wang, S., Li, Z.: A review on human activity recognition using vision-based method. J. Healthc. Eng. 2017, 31 (2017). https://doi.org/10.1155/2017/3090343.
Twomey, N., Diethe, T., Fafoutis, X., Elsts, A., McConville, R., Flach, P., Craddock, I.: A comprehensive study of activity recognition using accelerometers. Informatics 5(2), 1–37 (2018). https://doi.org/10.3390/informatics5020027; https://www.mdpi.com/2227-9709/5/2/27/htm
Wang, J., Chen, Y., Hao, S., Peng, X., Hu, L.: Deep learning for sensor-based activity recognition: A Survey. Pattern Recogn. Lett. 119(1), 3–11 (2019). https://doi.org/10.1016/j.patrec.2018.02.010; https://www.sciencedirect.com/science/article/abs/pii/S016786551830045X
Jobanputra, C., Bavishi, J., Doshi, N.: Human activity recognition: A survey. Procedia Comput. Sci. 155, 698–703 (2019). https://doi.org/10.1016/j.procs.2019.08.100; http://www.sciencedirect.com/science/article/pii/S1877050919310166
Lago, P., Alia, S.S., Takeda, S., Mairittha, T., Mairittha, N., Faiz, F., Nishimura, Y., Adachi, K., Okita, T., Charpillet, F., Inoue, S.: Nurse care activity recognition challenge: Summary and results. UbiComp/ISWC 2019 - Adjunct Proceedings of the 2019 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2019 ACM International Symposium on Wearable Computers, pp. 746–751 (2019). https://doi.org/10.1145/3341162.3345577; https://dl.acm.org/doi/10.1145/3341162.3345577
Martínez-Villaseñor, L., Ponce, H., Brieva, J., Moya-Albor, E., Núñez-Martínez, J., Peñafort-Asturiano, C.: Up-fall detection dataset: A multimodal approach. Sensors (Basel, Switzerland) 19(9) (2019). https://doi.org/10.3390/s19091988; https://www.mdpi.com/1424-8220/19/9/1988
Lago, P., Takeda, S., Adachi, K., Alia, S.S., Matsuki, M., Benaissa, B., Inoue, S., Charpillet, C.: Cooking activity dataset with macro and micro activities. IEEE DataPort (2020). https://doi.org/10.21227/hyzg-9m49
Lago, P., Takeda, S., Alia, S.S., Adachi, K., Benaissa, B., Charpillet, F., Inoue, S.: A dataset for complex activity recognition with micro and macro activities in a cooking scenario. Preprint (2020)
Alia, S.S., Lago, P., Takeda, S., Adachi, K., Benaissa, B., Ahad, M.A.R., Inoue, S.: Summary of the Cooking Activity Recognition Challenge. Human Activity Recognition Challenge, Smart Innovation, Systems and Technologies. Springer Nature (2020)
Eusha Kadir, M., Akash, P.S., Sharmin, S., Ali, A.A., Shoyaib, M.: Can a simple approach identify complex nurse care activity? UbiComp/ISWC 2019 - Adjunct Proceedings of the 2019 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2019 ACM International Symposium on Wearable Computers, pp. 736–740 (2019). https://doi.org/10.1145/3341162.3344859; https://dl.acm.org/doi/10.1145/3341162.3344859
Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q., Liu, T.Y.: LightGBM: A highly efficient gradient boosting decision tree. Adv. Neural Inform. Process. Syst. 30, 3146–3154 (2017). https://github.com/Microsoft/LightGBM
Gao, X., Luo, H.,Wang, Q., Zhao, F.,Ye, L., Zhang,Y.: A human activity recognition algorithm based on stacking denoising Autoencoder and LightGBM. Sensors (Basel, Switzerland) 19 (2019). https://doi.org/10.3390/s19040947
Akiba, T., Sano, S., Yanase, T., Ohta, T., Koyama, M.: Optuna: A next-generation hyperparameter optimization framework. In: KDD ’19: Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 2623–2631 (2019). https://doi.org/10.1145/3292500.3330701. https://dl.acm.org/doi/10.1145/3292500.3330701
Ziaeefard, M., Bergevin, R.: Semantic human activity recognition: A literature review. Pattern Recogn. 48(8), 2329–2345 (2015). https://doi.org/10.1016/j.patcog.2015.03.006. https://www.sciencedirect.com/science/article/abs/pii/S0031320315000953
Ramasamy Ramamurthy, S., Roy, N.: Recent trends in machine learning for human activity recognition - A survey. WIREs Data Min. Knowl. Dis. 8, e1254 (2018). https://doi.org/10.1002/widm.1254.
Ye, J., Dobson, S., Zambonelli, F.: Lifelong learning in sensor-based human activity recognition. IEEE Pervasive Comput.18(3), 49–58 (2019). https://doi.org/10.1109/MPRV.2019.2913933. https://ieeexplore.ieee.org/document/8903481
Ordóñez, F.J., Roggen, D.: Deep convolutional and LSTM recurrent neural networks for multimodal wearable activity recognition. Sensors 16(1), 115 (2016). https://doi.org/10.3390/s16010115. https://www.mdpi.com/1424-8220/16/1/115
Yao, S., Hu, S., Zhao, Y., Zhang, A., Abdelzaher, T.: DeepSense: A unified deep learning framework for time-series mobile sensing data processing. Proceedings of the 26th International World Wide Web Conference, pp. 351–360 (2017). https://doi.org/10.1145/3038912.3052577. https://dl.acm.org/doi/10.1145/3038912.3052577
Ma, H., Li, W., Zhang, X., Gao, S., Lu, S.: AttnSense: Multi-level attention mechanism for multimodal human activity recognition. Proceedings of the 28th International Joint Conference on Artificial Intelligence, pp. 3109–3115 (2019). https://doi.org/10.24963/ijcai.2019/431. https://www.ijcai.org/Proceedings/2019/431
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Appendix
Appendix
We summarize in Table 4 the proposed method we elucidated in this paper, as well as the experiment environment in which our method was carried out.
Rights and permissions
Copyright information
© 2021 The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this chapter
Cite this chapter
Kojima, R., Legaspi, R., Yoshihara, K., Wada, S. (2021). Let’s Not Make It Complicated—Using Only LightGBM and Naive Bayes for Macro- and Micro-Activity Recognition from a Small Dataset. In: Ahad, M.A.R., Lago, P., Inoue, S. (eds) Human Activity Recognition Challenge. Smart Innovation, Systems and Technologies, vol 199. Springer, Singapore. https://doi.org/10.1007/978-981-15-8269-1_3
Download citation
DOI: https://doi.org/10.1007/978-981-15-8269-1_3
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-15-8268-4
Online ISBN: 978-981-15-8269-1
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)