Fusion Models for Generalized Classification of Multi-Axial Human Movement: Validation in Sport Performance
We introduce a set of input models for fusing information from ensembles of wearable sensors supporting human performance and telemedicine. Veracity is demonstrated in action classification related to sport, specifically strikes in boxing and taekwondo. Four input models, formulated to be compatible...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-12-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/24/8409 |
| _version_ | 1797500792368791552 |
|---|---|
| author | Rajesh Amerineni Lalit Gupta Nathan Steadman Keshwyn Annauth Charles Burr Samuel Wilson Payam Barnaghi Ravi Vaidyanathan |
| author_facet | Rajesh Amerineni Lalit Gupta Nathan Steadman Keshwyn Annauth Charles Burr Samuel Wilson Payam Barnaghi Ravi Vaidyanathan |
| author_sort | Rajesh Amerineni |
| collection | DOAJ |
| description | We introduce a set of input models for fusing information from ensembles of wearable sensors supporting human performance and telemedicine. Veracity is demonstrated in action classification related to sport, specifically strikes in boxing and taekwondo. Four input models, formulated to be compatible with a broad range of classifiers, are introduced and two diverse classifiers, dynamic time warping (DTW) and convolutional neural networks (CNNs) are implemented in conjunction with the input models. Seven classification models fusing information at the input-level, output-level, and a combination of both are formulated. Action classification for 18 boxing punches and 24 taekwondo kicks demonstrate our fusion classifiers outperform the best DTW and CNN uni-axial classifiers. Furthermore, although DTW is ostensibly an ideal choice for human movements experiencing non-linear variations, our results demonstrate deep learning fusion classifiers outperform DTW. This is a novel finding given that CNNs are normally designed for multi-dimensional data and do not specifically compensate for non-linear variations within signal classes. The generalized formulation enables subject-specific movement classification in a feature-blind fashion with trivial computational expense for trained CNNs. A commercial boxing system, ‘Corner’, has been produced for real-world mass-market use based on this investigation providing a basis for future telemedicine translation. |
| first_indexed | 2024-03-10T03:08:56Z |
| format | Article |
| id | doaj.art-a6c01bd93ac0477b9fbdb90fcce12b47 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T03:08:56Z |
| publishDate | 2021-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-a6c01bd93ac0477b9fbdb90fcce12b472023-11-23T10:31:04ZengMDPI AGSensors1424-82202021-12-012124840910.3390/s21248409Fusion Models for Generalized Classification of Multi-Axial Human Movement: Validation in Sport PerformanceRajesh Amerineni0Lalit Gupta1Nathan Steadman2Keshwyn Annauth3Charles Burr4Samuel Wilson5Payam Barnaghi6Ravi Vaidyanathan7Department of Electrical Engineering, Southern Illinois University, Carbondale, IL 62901, USADepartment of Electrical Engineering, Southern Illinois University, Carbondale, IL 62901, USADepartment of Mechanical Engineering, Imperial College London, London SW7 2AZ, UKDepartment of Mechanical Engineering, Imperial College London, London SW7 2AZ, UKDepartment of Mechanical Engineering, Imperial College London, London SW7 2AZ, UKDepartment of Mechanical Engineering, Imperial College London, London SW7 2AZ, UKDepartment of Brain Sciences, Imperial College London, London W12 0NN, UKDepartment of Mechanical Engineering, Imperial College London, London SW7 2AZ, UKWe introduce a set of input models for fusing information from ensembles of wearable sensors supporting human performance and telemedicine. Veracity is demonstrated in action classification related to sport, specifically strikes in boxing and taekwondo. Four input models, formulated to be compatible with a broad range of classifiers, are introduced and two diverse classifiers, dynamic time warping (DTW) and convolutional neural networks (CNNs) are implemented in conjunction with the input models. Seven classification models fusing information at the input-level, output-level, and a combination of both are formulated. Action classification for 18 boxing punches and 24 taekwondo kicks demonstrate our fusion classifiers outperform the best DTW and CNN uni-axial classifiers. Furthermore, although DTW is ostensibly an ideal choice for human movements experiencing non-linear variations, our results demonstrate deep learning fusion classifiers outperform DTW. This is a novel finding given that CNNs are normally designed for multi-dimensional data and do not specifically compensate for non-linear variations within signal classes. The generalized formulation enables subject-specific movement classification in a feature-blind fashion with trivial computational expense for trained CNNs. A commercial boxing system, ‘Corner’, has been produced for real-world mass-market use based on this investigation providing a basis for future telemedicine translation.https://www.mdpi.com/1424-8220/21/24/8409sports biomechanicshuman performancemotion trackingwearable sensorsIMUssensor fusion |
| spellingShingle | Rajesh Amerineni Lalit Gupta Nathan Steadman Keshwyn Annauth Charles Burr Samuel Wilson Payam Barnaghi Ravi Vaidyanathan Fusion Models for Generalized Classification of Multi-Axial Human Movement: Validation in Sport Performance Sensors sports biomechanics human performance motion tracking wearable sensors IMUs sensor fusion |
| title | Fusion Models for Generalized Classification of Multi-Axial Human Movement: Validation in Sport Performance |
| title_full | Fusion Models for Generalized Classification of Multi-Axial Human Movement: Validation in Sport Performance |
| title_fullStr | Fusion Models for Generalized Classification of Multi-Axial Human Movement: Validation in Sport Performance |
| title_full_unstemmed | Fusion Models for Generalized Classification of Multi-Axial Human Movement: Validation in Sport Performance |
| title_short | Fusion Models for Generalized Classification of Multi-Axial Human Movement: Validation in Sport Performance |
| title_sort | fusion models for generalized classification of multi axial human movement validation in sport performance |
| topic | sports biomechanics human performance motion tracking wearable sensors IMUs sensor fusion |
| url | https://www.mdpi.com/1424-8220/21/24/8409 |
| work_keys_str_mv | AT rajeshamerineni fusionmodelsforgeneralizedclassificationofmultiaxialhumanmovementvalidationinsportperformance AT lalitgupta fusionmodelsforgeneralizedclassificationofmultiaxialhumanmovementvalidationinsportperformance AT nathansteadman fusionmodelsforgeneralizedclassificationofmultiaxialhumanmovementvalidationinsportperformance AT keshwynannauth fusionmodelsforgeneralizedclassificationofmultiaxialhumanmovementvalidationinsportperformance AT charlesburr fusionmodelsforgeneralizedclassificationofmultiaxialhumanmovementvalidationinsportperformance AT samuelwilson fusionmodelsforgeneralizedclassificationofmultiaxialhumanmovementvalidationinsportperformance AT payambarnaghi fusionmodelsforgeneralizedclassificationofmultiaxialhumanmovementvalidationinsportperformance AT ravivaidyanathan fusionmodelsforgeneralizedclassificationofmultiaxialhumanmovementvalidationinsportperformance |