Reconocimiento de actividades humanas utilizando máquinas de soporte vectorial penalizadas y modelos ocultos de Markov

Autores/as

DOI:

https://doi.org/10.17533/udea.redin.20210532

Palabras clave:

multi sensor, fusión de datos, movimientos primitivos, aprendizaje de máquinas, datos traslapados

Resumen

La detección de actividades humanas ha logrado evolucionar debido a los avances y desarrollos de técnicas del aprendizaje de máquinas, las cuales han permitido dar soluciones a nuevos desafíos sin ignorar las dificultades que aún persisten y abogan atención; uno de ellos concierne a la sensibilidad que presenta el modelo de aprendizaje ante información traslapada, desbalanceada y atípica que repercute propiamente en el desempeño del modelo. En este artículo se evalúa una metodología para la clasificación de actividades humanas que castiga información con imperfectos. El proceso metodológico se lleva cabo por medio de dos clasificadores redundantes, una Máquinas de Vectores de Soporte penalizada que detecta los sub movimientos (micromovimientos) y luego un Modelo Oculto de Markov que predice la actividad dada la secuencia de micro movimiento. El desempeño del método fue comparado con técnicas del estado de arte, los resultados sugieren un avance significativo en la detección de micromovientos frente a los obtenidos con paradigmas no penalizados. En este trabajo se obtiene un adecuado desempeño en la clasificación de movimientos primitivos, con aciertos del 95,15% para el Kinect One®, del 96,86% para la red de sensores IMU y del 67,51% para la red de sensores EMG. Lo anterior impacta directamente la detección de actividades físicas con aciertos mayores al 95% de eficiencia.

|Resumen
= 1327 veces | PDF (ENGLISH)
= 608 veces|

Descargas

Los datos de descargas todavía no están disponibles.

Biografía del autor/a

Leidy Esperanza Pamplona-Berón, Universidad Santiago de Cali

Magister en Ingeniería Eléctrica.

Carlos Alberto Henao Baena, SENA

Tecnoparque- Nodo Pereira, Escuela de Telecomunicaciones y Electrónica.

Andrés Felipe Calvo-Salcedo, Universidad Tecnológica de Pereira

Magister Ingeniería Eléctrica.

Citas

L. Cao, Y. Wang, B. Zhang, Q. Jin, and A. V. Vasilakos, “Gchar: An efficient group-based context aware human activity recognition on smartphone,” Journal of Parallel and Distributed Computing, vol. 118, Part 1, Aug. 2018. [Online]. Available: https://doi.org/10.1016/j.jpdc.2017.05.007

A. Khan, N. Hammerla, S. Mellor, and T. Plötz, “Optimising sampling rates for accelerometer-based human activity recognition,” Pattern Recognition Letters, vol. 73, Apr. 1, 2016. [Online]. Available: https://doi.org/10.1016/j.patrec.2016.01.001

R. Gravina, P. Alinia, H. Ghasemzadeh, and G. Fortino, “Multi-sensor fusion in body sensor networks: State-of-the-art and research challenges,” Information Fusion, vol. 35, May. 2017. [Online]. Available: https://doi.org/10.1016/j.inffus.2016.09.005

Y. L. Chen and et al, “Dimensionality reduction of data sequences for human activity recognition,” Neurocomputing, vol. 210, Oct. 19, 2016. [Online]. Available: https://doi.org/10.1016/j.neucom.2015.11.126

W. Takano, H. Imagawa, and Y. Nakamura, “Spatio-temporal structure of human motion primitives and its application to motion prediction,” Robotics and Autonomous Systems, vol. 75, Part B, Jan. 2016. [Online]. Available: https://doi.org/10.1016/j.robot.2015.09.017

S. Morales, “Identificación de actividad humana usando aprendizaje no supervisado en sistemas multimodales,” M.S. thesis, Facultad de Ingenierías, Universidad Tecnológica de Pereira, Pereira, CO, 2016.

A. F. Calvo, “Reconocimiento automático de actividades físicas humanas en sistemas multimodales,” M.S. thesis, Facultad de Ingenierías, Universidad Tecnológica de Pereira, Pereira, CO, 2015.

M. Jiang, J. Kong, G. Bebis, and H. Huo, “Informative joints based human action recognition using skeleton contexts,” Signal Processing: Image Communication, vol. 33, Apr. 2015. [Online]. Available: https://doi.org/10.1016/j.image.2015.02.004

R. Qiao, L. Liu, C. Shen, and A. V. Den, “Learning discriminative trajectorylet detector sets for accurate skeleton-based action recognition,” Pattern Recognition, vol. 66, Jun. 2017. [Online]. Available: https://doi.org/10.1016/j.patcog.2017.01.015

A. Bayat, M. Pomplun, and D. Tran, “A study on human activity recognition using accelerometer data from smartphones,” Procedia Computer Science, vol. 34, 2014. [Online]. Available: https://doi.org/10.1016/j.procs.2014.07.009

A. Akbari, X. Thomas, and R. Jafari, “Automatic noise estimation and context-enhanced data fusion of imu and kinect for human motion measurement,” in 2017 IEEE 14thInternational Conference on Wearable and Implantable Body Sensor Networks (BSN), Eindhoven, NL, 2017.

I. Serrano, V. Kyrki, D. Kragic, and M. Larsson, “Action recognition and understanding through motor primitives,” Advanced Robotics, vol. 21, no. 15, Nov. 2007. [Online]. Available: https://doi.org/10.1163/156855307782506156

S. Gaglio, G. L. Re, and M. Morana, “Human activity recognition process using 3-D posture data,” IEEE Transactions on Human-Machine Systems, vol. 45, no. 5, Dec. 18, 2014. [Online]. Available: https://doi.org/10.1109/THMS.2014.2377111

M. Zhang and A. Sawchuk, “Motion primitive-based human activity recognition using a bag-of-features approach,” in IHI ’12: Proceedings of the 2nd ACM SIGHIT International Health Informatics Symposium, 2012, pp. 631––640.

J. F. S. Lin, M. Karg, and D. Kulić, “Movement primitive segmentation for human motion modeling: A framework for analysis,” IEEE Transactions on Human-Machine Systems, vol. 46, no. 3, Jun. 2016. [Online]. Available: https://doi.org/10.1109/THMS.2015.2493536

A. F. Calvo, G. A. Holguin, and H. Medeiros, “Human activity recognition using multi-modal data fusion,” in CIARP 2018: Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications, 2018, pp. 946–953.

B. Wang, C. Yang, and Q. Xie, “Human-machine interfaces based on emg and kinect applied to teleoperation of a mobile humanoid robot,” in Proceedings of the 10th World Congress on Intelligent Control and Automation, Beijing, CN, 2012, pp. 3903–3908.

S. Feng and R. Murray, “Fusing kinect sensor and inertial sensors with multi-rate kalman filter,” in IET Conference on Data Fusion & Target Tracking 2014: Algorithms and Applications (DF&TT 2014), Liverpool, UK, 2014, pp. 1–8.

C. Xiang, H. H. Hsu, W. Y. Hwang, and J. Ma, “Comparing real-time human motion capture system using inertial sensors with microsoft kinect,” in 2014 7th International Conference on Ubi-Media Computing and Workshops, Ulaanbaatar, MN, 2014, pp. 53–58.

M. Caon and et al, “Kinesiologic electromyography for activity recognition,” in PETRA ’13: Proceedings of the 6th International Conference on PErvasive Technologies Related to Assistive Environments, New York, USA, 2013, pp. 1–7.

H. Koskimaki and P. Siirtola, “Accelerometer vs. electromyogram in activity recognition,” ADCAIJ: Advances in Distributed Computing and Artificial Intelligence Journal, vol. 5, no. 3, Nov. 2016. [Online]. Available: https://doi.org/10.14201/ADCAIJ2016533142

A. Castellani, D. Botturi, M. Bicego, and P. Fiorini, “Hybrid hmm/svm model for the analysis and segmentation of teleoperation tasks,” in IEEE International Conference on Robotics and Automation, 2004, New Orleans, LA, USA, 2004, pp. 2918–2923.

B. Hannaford and P. Lee, “Hidden markov model analysis of force/torque information in telemanipulation,” The International journal of robotics research, vol. 10, no. 5, Oct. 1, 1991. [Online]. Available: https://doi.org/10.1177/027836499101000508

K. Chen and et al, “Deep learning for sensor-based human activity recognition: Overview, challenges and opportunities,” J. ACM, vol. 37, no. 4, Aug. 2018. [Online]. Available: https://doi.org/10.1145/3447744

F. Ofli, R. Chaudhry, G. Kurillo, R. Vidal, and R. Bajcsy, “Berkeley mhad: A comprehensive multimodal human action database,” in 2013 IEEE Workshop on Applications of Computer Vision (WACV), Tampa, FL, USA, 2013, pp. 53–60.

H. H. Pham, L. Khoudour, A. Crouzil, P. Zegers, and S. A. Velastin, “Exploiting deep residual networks for human action recognition from skeletal data,” Computer Vision and Image Understanding, vol. 170, May. 2018. [Online]. Available: https://doi.org/10.1016/j.cviu.2018.03.003

C. Roldán, “Estudio de la cinemática del miembro superior e inferior mediante sensores inerciales,” Ph. D. dissertation, Facultad de Ciencias de la Salud, Universidad de Málaga, Málaga, ES, 2017.

D. A. Winter, “Moments of force and mechanical power in jogging,” Journal of biomechanics, vol. 16, no. 1, 1983. [Online]. Available: https://doi.org/10.1016/0021-9290(83)90050-7

S. A. Dugan and K. P. Bhat, “Biomechanics and analysis of running gait,” Physical Medicine and Rehabilitation Clinics, vol. 16, no. 3, Aug. 2005. [Online]. Available: https://doi.org/10.1016/j.pmr.2005.02.007

A. Gomez, R. Becerro, and M. E. Losa, “Reliability of the optogait portable photoelectric cell system for the quantification of spatial-temporal parameters of gait in young adults,” Gait & posture, vol. 50, Oct. 2016. [Online]. Available: https://doi.org/10.1016/j.gaitpost.2016.08.035

S. Zennaro and et al, “Performance evaluation of the 1st and 2 nd generation kinect for multimedia applications,” in 2015 IEEE International Conference on Multimedia and Expo (ICME), Turin, IT, 2015, pp. 1–6.

D. Pagliari and L. Pinto, “Calibration of kinect for xbox one and comparison between the two generations of microsoft sensors,” Sensors, vol. 15, no. 11, Oct. 2015. [Online]. Available: https://doi.org/10.3390/s151127569

H. Wu, W. Pan, X. Xiong, and S. Xu, “Human activity recognition based on the combined SVM&HMM,” in 2014 IEEE International Conference on Information and Automation (ICIA), Hailar, CN, 2014, pp. 219–224.

M. Zhang and A. A. Sawchuk, “A feature selection-based framework for human activity recognition using wearable multimodal sensors,” in BodyNets ’11: Proceedings of the 6thInternational Conference on Body Area Networks, Zadar, HR, 2011, pp. 92–98.

B. Schölkopf and A. J. Smola, Learning with Kernels – Support Vector Machines, Regularization, Optimization and Beyond, 1st ed. Cambridge, MA, USA: MIT Press, 2001.

S. Shao, K. Shen, C. J. Ong, E. P. V. Wilder, and X. Li, “Automatic EEG artifact removal: A weighted support vector machine approach with error correction,” IEEE Transactions on Biomedical Engineering, vol. 56, no. 2, Feb. 2009. [Online]. Available: https://doi.org/10.1109/TBME.2008.2005969

J. F. Gallego and D. F. Rengifo, “Comparación de técnicas de reducción de dimensionalidad para la clasificación de actividades ffiısicas humanas utilizando métodos estadísticos,” Undergraduate degree, Facultad de Ingeniería, Universidad Tecnológica de Pereira, Pereira, CO, 2016.

L. Jiang and R. Yao, “Modelling personal thermal sensations using c-support vector classification (c-svc) algorithm,” Building and Environment, vol. 99, Apr. 2016. [Online]. Available: https://doi.org/10.1016/j.buildenv.2016.01.022

N. Becker, W. Werft, G. Toedt, P. Lichter, and A. Benner, “penalizedsvm: a r-package for feature selection svm classification,” Bioinformatics, vol. 25, no. 13, Jul. 1, 2009. [Online]. Available: https://doi.org/10.1093/bioinformatics/btp286

Descargas

Publicado

2021-05-24

Cómo citar

Pamplona-Berón, L. E., Henao Baena, C. A., & Calvo-Salcedo, A. F. (2021). Reconocimiento de actividades humanas utilizando máquinas de soporte vectorial penalizadas y modelos ocultos de Markov. Revista Facultad De Ingeniería Universidad De Antioquia, (103), 152–163. https://doi.org/10.17533/udea.redin.20210532