Human activity recognition using penalized support vector machines and hidden Markov models.
Fuente:
Redalyc
Saved in:
| Main Author: | |
|---|---|
| Format: | Artículo científico |
| Language: | en |
| Published: |
Universidad de Antioquia
2022
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1876429040292724736 |
|---|---|
| author | Leidy Esperanza Pamplona Berón |
| author_facet | Leidy Esperanza Pamplona Berón |
| contents | Human activity recognition using penalized support vector machines and hidden Markov models. Leidy Esperanza Pamplona Berón Carlos Alberto Henao Baena Andrés Felipe Calvo Salcedo Ingeniería Multisensor data fusion primitive motion machine learning overlapping databases Human activity detection has evolved due to the advances and developments of machine learning techniques, which have enabled solutions to new challenges without ignoring prevalent difficulties that need to be addressed. One of the challenges is the learning model’s sensitivity regarding the unbalanced, atypical, and overlapping information that directly affects the performance of the model. This article evaluates a methodology for the classification of human activities that penalizes defective information. The methodology is carried out through two redundant classifiers, a penalized support vector machine that detects the sub-movements (micro-movements) and the Marvok Hidden Model that predicts the activity given the micro- movements sequence. The performance of the method was compared with state-of-the-art techniques, and the findings suggested significative advance in the detection of micro-movements compared to the data obtained with non-penalized paradigms. In this research, an adequate performance is found in the classification of primitive movements, with hit rates of 95.15% for the Kinect One®, 96.86% for the IMU sensor network, and 67.51% for the EMG sensor network. 2022 artículo científico 0120-6230 https://www.redalyc.org/articulo.oa?id=43070264014 https://www.redalyc.org/journal/430/43070264014/ https://www.redalyc.org/journal/430/43070264014/html/ https://www.redalyc.org/journal/430/43070264014/43070264014.epub https://www.redalyc.org/journal/430/43070264014/movil 10.17533/udea.redin.20210532 en http://www.redalyc.org/revista.oa?id=430 Revista Facultad de Ingeniería Universidad de Antioquia application/pdf Universidad de Antioquia Revista Facultad de Ingeniería Universidad de Antioquia (Colombia) Num.103 |
| format | Artículo científico |
| id | redalyc_43070264014 |
| institution | Redalyc |
| language | en |
| publishDate | 2022 |
| publisher | Universidad de Antioquia |
| spellingShingle | Human activity recognition using penalized support vector machines and hidden Markov models. Leidy Esperanza Pamplona Berón Ingeniería Multisensor data fusion primitive motion machine learning overlapping databases Human activity recognition using penalized support vector machines and hidden Markov models. Leidy Esperanza Pamplona Berón Carlos Alberto Henao Baena Andrés Felipe Calvo Salcedo Ingeniería Multisensor data fusion primitive motion machine learning overlapping databases Human activity detection has evolved due to the advances and developments of machine learning techniques, which have enabled solutions to new challenges without ignoring prevalent difficulties that need to be addressed. One of the challenges is the learning model’s sensitivity regarding the unbalanced, atypical, and overlapping information that directly affects the performance of the model. This article evaluates a methodology for the classification of human activities that penalizes defective information. The methodology is carried out through two redundant classifiers, a penalized support vector machine that detects the sub-movements (micro-movements) and the Marvok Hidden Model that predicts the activity given the micro- movements sequence. The performance of the method was compared with state-of-the-art techniques, and the findings suggested significative advance in the detection of micro-movements compared to the data obtained with non-penalized paradigms. In this research, an adequate performance is found in the classification of primitive movements, with hit rates of 95.15% for the Kinect One®, 96.86% for the IMU sensor network, and 67.51% for the EMG sensor network. 2022 artículo científico 0120-6230 https://www.redalyc.org/articulo.oa?id=43070264014 https://www.redalyc.org/journal/430/43070264014/ https://www.redalyc.org/journal/430/43070264014/html/ https://www.redalyc.org/journal/430/43070264014/43070264014.epub https://www.redalyc.org/journal/430/43070264014/movil 10.17533/udea.redin.20210532 en http://www.redalyc.org/revista.oa?id=430 Revista Facultad de Ingeniería Universidad de Antioquia application/pdf Universidad de Antioquia Revista Facultad de Ingeniería Universidad de Antioquia (Colombia) Num.103 |
| title | Human activity recognition using penalized support vector machines and hidden Markov models. |
| topic | Ingeniería Multisensor data fusion primitive motion machine learning overlapping databases |
| url | https://www.redalyc.org/articulo.oa?id=43070264014 https://www.redalyc.org/journal/430/43070264014/ https://www.redalyc.org/journal/430/43070264014/html/ https://www.redalyc.org/journal/430/43070264014/43070264014.epub https://www.redalyc.org/journal/430/43070264014/movil |