Abstract
In this paper we leverage voting rules as an aggregation technique in classifier combination. We propose a voting rules-based algorithm to improve the recognition performance of human action recognition on raw depth maps. The recognition is performed by a classifier committee that is built on deep embedding features. It consists of softmax logistic regression classifiers delivering well calibrated outputs. They have been trained as one-vs-all on the concatenated class-specific features and features common for all classes. Various ranked voting rules as the aggregation technique have been investigated in decision making. The results achieved by such algorithms were compared with results achieved in optimized weighted voting. Differential evolution and sequential least squares programming have been selected for optimized weighted voting. We demonstrate experimentally that both optimized weighted voting and voting rules improve results achieved by soft voting. The best results have been achieved using voting rules determined by Coombs algorithm. We demonstrate experimentally that on SYSU 3DHOI dataset the proposed algorithm outperforms by a large margin all recent algorithms.
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
Polikar, R.: Ensemble based systems in decision making. IEEE Circuits Syst. Mag. 6(3), 21–45 (2006)
Kuncheva, L.I., Whitaker, C.J.: Measures of diversity in classifier ensembles and their relationship with the ensemble accuracy. Mach. Learn. 51(2), 181–207 (2003)
Xu, L., Krzyzak, A., Suen, C.: Methods of combining multiple classifiers and their applications to handwriting recognition. IEEE Trans. SMC 22(3), 418–435 (1992)
Kuncheva, L.I.: A theoretical study on six classifier fusion strategies. IEEE Trans. Pattern Anal. Mach. Intell. 24(2), 281–286 (2002)
Fumera, G., Roli, F.: A theoretical and experimental analysis of linear combiners for multiple classifier systems. IEEE Trans. PAMI 27(6), 942–956 (2005)
Bonab, H., Can, F.: Less is more: A comprehensive framework for the number of components of ensemble classifiers. IEEE Trans. on Neural Networks and Learning Systems 30(9) (2019) 2735–2745
van Erp, M., Vuurpijl, L., Schomaker, L.: An overview and comparison of voting methods for pattern recognition. In: Proceedings of Eighth International Workshop on Frontiers in Handwriting Recognition, pp. 195–200 (2002)
Liang, B., Zheng, L.: A survey on human action recognition using depth sensors. In: Int. Conf. on Digital Image Comp.: Techn. and Appl., pp. 1–8(2015)
Wang, L., Huynh, D.Q., Koniusz, P.: A comparative review of recent Kinect-based action recognition algorithms. IEEE Trans. Image Process. 29, 15–28 (2020)
Haque, A., Milstein, A., Fei-Fei, L.: Illuminating the dark spaces of healthcare with ambient intelligence. Nature 585(7824), 193–202 (2020)
Ren, B., Liu, M., Ding, R., Liu, H.: A survey on 3D skeleton-based action recognition using learning method. arXiv, 2002.05907 (2020)
Yang, X., Zhang, C., Tian, Y.L.: Recognizing actions using depth motion maps-based histograms of oriented gradients. In: Proceedings of the 20th ACM International Conference on Multimedia, pp. 1057–1060. ACM (2012)
Xia, L., Aggarwal, J.: Spatio-temporal depth cuboid similarity feature for activity recognition using depth camera. In: CVPR, pp. 2834–2841(2013)
Li, C., Huang, Q., Li, X., Wu, Q.: A multi-scale human action recognition method based on Laplacian pyramid depth motion images. In: Proceedings the 2nd ACM International Conference on Multimedia in Asia. ACM (2021)
Majumder, S., Kehtarnavaz, N.: Vision and inertial sensing fusion for human action recognition: a review. IEEE Sensors J. 21(3), 2454–2467 (2021)
Trelinski, J., Kwolek, B.: Deep embedding features for action recognition on raw depth maps. In: Paszynski, M., Kranzlmüller, D., Krzhizhanovskaya, V.V., Dongarra, J.J., Sloot, P.M.A. (eds.) ICCS 2021. LNCS, vol. 12744, pp. 95–108. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-77967-2_9
Hassan, S., Hemeida, A.M., Alkhalaf, S., Mohamed, A.A., Senjyu, T.: Multi-variant differential evolution algorithm for feature selection. Scientific Reports 10(1), October 2020
Erdélyi, G., Fellows, M.R., Rothe, J., Schend, L.: Control complexity in Bucklin and fallback voting: a theoretical analysis. J. Comput. Syst. Sci. 81(4), 632–660 (2015)
Pacuit, E.: Voting methods. Stanford Encyclopedia of Philosophy (Fall 2017 Ed.), Edward N. Zalta (ed.) (2017)
Paliwal, K., Agarwal, A., Sinha, S.: A modification over Sakoe and Chiba’s dynamic time warping algorithm for isolated word recognition. Signal Proc. 4(4), 329–333 (1982)
Meert, W., Hendrickx, K., Craenendonck, T.V.: DTAIdistance, ver. 2.0 (2021). https://zenodo.org/record/3981067
Hu, J., Zheng, W., Lai, J., Zhang, J.: Jointly learning heterogeneous features for RGB-D activity recognition. In: CVPR, pp. 5344–5352 (2015)
Hu, J., Zheng, W., Ma, L., Wang, G., Lai, J., Zhang, J.: Early action prediction by soft regression. IEEE Trans. PAMI 41(11), 2568–2583 (2019)
Wang, X., Hu, J.F., Lai, J.H., Zhang, J., Zheng, W.S.: Progressive teacher-student learning for early action prediction. In: CVPR, pp. 3551–3560 (2019)
Zhang, P., Lan, C., Zeng, W., Xing, J., Xue, J., Zheng, N.: Semantics-guided neural networks for efficient skeleton-based human action recognition. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 1109–1118. IEEE
Ke, Q., Bennamoun, M., Rahmani, H., An, S., Sohel, F., Boussaid, F.: Learning latent global network for skeleton-based action prediction. IEEE Trans. Img. Proc. 29, 959–970 (2020)
Hu, J.F., Zheng, W.S., Ma, L., Wang, G., Lai, J.: Real-time RGB-D activityprediction by soft regression. In: European Conf. on Comp. Vision, pp. 280–296. Springer (2016)
Ren, Z., Zhang, Q., Gao, X., Hao, P., Cheng, J.: Multi-modality learning for human action recognition. Multimed. Tools Appl. 80(11), 16185–16203 (2020). https://doi.org/10.1007/s11042-019-08576-z
Acknowledgment
This work was supported by Polish National Science Center (NCN) under a research grant 2017/27/B/ST6/01743.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Treliński, J., Kwolek, B. (2021). Decision Combination in Classifier Committee Built on Deep Embedding Features. In: Nguyen, N.T., Iliadis, L., Maglogiannis, I., Trawiński, B. (eds) Computational Collective Intelligence. ICCCI 2021. Lecture Notes in Computer Science(), vol 12876. Springer, Cham. https://doi.org/10.1007/978-3-030-88081-1_36
Download citation
DOI: https://doi.org/10.1007/978-3-030-88081-1_36
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-88080-4
Online ISBN: 978-3-030-88081-1
eBook Packages: Computer ScienceComputer Science (R0)