SCB-YOLOv5: a lightweight intelligent detection model for athletes’ normative movements
Abstract Intelligent detection of athlete behavior is beneficial for guiding sports instruction. Existing mature target detection algorithms provide significant support for this task. However, large-scale target detection algorithms often encounter more challenges in practical application scenarios....
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-04-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-024-59218-w |
| _version_ | 1797199559545323520 |
|---|---|
| author | Qing Du Lian Tang Ya Li |
| author_facet | Qing Du Lian Tang Ya Li |
| author_sort | Qing Du |
| collection | DOAJ |
| description | Abstract Intelligent detection of athlete behavior is beneficial for guiding sports instruction. Existing mature target detection algorithms provide significant support for this task. However, large-scale target detection algorithms often encounter more challenges in practical application scenarios. We propose SCB-YOLOv5, to detect standardized movements of gymnasts. First, the movements of aerobics athletes were captured, labeled using the labelImg software, and utilized to establish the athlete normative behavior dataset, which was then enhanced by the dataset augmentation using Mosaic9. Then, we improved the YOLOv5 by (1) incorporating the structures of ShuffleNet V2 and convolutional block attention module to reconstruct the Backbone, effectively reducing the parameter size while maintaining network feature extraction capability; (2) adding a weighted bidirectional feature pyramid network into the multiscale feature fusion, to acquire precise channel and positional information through the global receptive field of feature maps. Finally, SCB-YOLOv5 was lighter by 56.9% than YOLOv5. The detection precision is 93.7%, with a recall of 99% and mAP value of 94.23%. This represents a 3.53% improvement compared to the original algorithm. Extensive experiments have verified that our method. SCB-YOLOv5 can meet the requirements for on-site athlete action detection. Our code and models are available at https://github.com/qingDu1/SCB-YOLOv5 . |
| first_indexed | 2024-04-24T07:17:41Z |
| format | Article |
| id | doaj.art-a267dbac69604eb1bfc76fde2a684687 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-24T07:17:41Z |
| publishDate | 2024-04-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-a267dbac69604eb1bfc76fde2a6846872024-04-21T11:14:20ZengNature PortfolioScientific Reports2045-23222024-04-0114111110.1038/s41598-024-59218-wSCB-YOLOv5: a lightweight intelligent detection model for athletes’ normative movementsQing Du0Lian Tang1Ya Li2School of Resource Environment and Safety Engineering, University of South ChinaSchool of Sports Science and Engineering, Hunan Institute of EngineeringSchool of Electrical Information Engineering, Hunan Institute of EngineeringAbstract Intelligent detection of athlete behavior is beneficial for guiding sports instruction. Existing mature target detection algorithms provide significant support for this task. However, large-scale target detection algorithms often encounter more challenges in practical application scenarios. We propose SCB-YOLOv5, to detect standardized movements of gymnasts. First, the movements of aerobics athletes were captured, labeled using the labelImg software, and utilized to establish the athlete normative behavior dataset, which was then enhanced by the dataset augmentation using Mosaic9. Then, we improved the YOLOv5 by (1) incorporating the structures of ShuffleNet V2 and convolutional block attention module to reconstruct the Backbone, effectively reducing the parameter size while maintaining network feature extraction capability; (2) adding a weighted bidirectional feature pyramid network into the multiscale feature fusion, to acquire precise channel and positional information through the global receptive field of feature maps. Finally, SCB-YOLOv5 was lighter by 56.9% than YOLOv5. The detection precision is 93.7%, with a recall of 99% and mAP value of 94.23%. This represents a 3.53% improvement compared to the original algorithm. Extensive experiments have verified that our method. SCB-YOLOv5 can meet the requirements for on-site athlete action detection. Our code and models are available at https://github.com/qingDu1/SCB-YOLOv5 .https://doi.org/10.1038/s41598-024-59218-wDigital sportsAction recognitionDeep learningAttention module |
| spellingShingle | Qing Du Lian Tang Ya Li SCB-YOLOv5: a lightweight intelligent detection model for athletes’ normative movements Scientific Reports Digital sports Action recognition Deep learning Attention module |
| title | SCB-YOLOv5: a lightweight intelligent detection model for athletes’ normative movements |
| title_full | SCB-YOLOv5: a lightweight intelligent detection model for athletes’ normative movements |
| title_fullStr | SCB-YOLOv5: a lightweight intelligent detection model for athletes’ normative movements |
| title_full_unstemmed | SCB-YOLOv5: a lightweight intelligent detection model for athletes’ normative movements |
| title_short | SCB-YOLOv5: a lightweight intelligent detection model for athletes’ normative movements |
| title_sort | scb yolov5 a lightweight intelligent detection model for athletes normative movements |
| topic | Digital sports Action recognition Deep learning Attention module |
| url | https://doi.org/10.1038/s41598-024-59218-w |
| work_keys_str_mv | AT qingdu scbyolov5alightweightintelligentdetectionmodelforathletesnormativemovements AT liantang scbyolov5alightweightintelligentdetectionmodelforathletesnormativemovements AT yali scbyolov5alightweightintelligentdetectionmodelforathletesnormativemovements |