Fine-grained action recognition of boxing punches from depth imagery

Kasiri, Soudeh; Fookes, Clinton; Sridharan, Sridha; Morgan, Stuart

doi:10.1016/j.cviu.2017.04.007

Cited by 25 publications

(22 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Both machine (n = 16) and deep learning (n = 7) algorithms were used to recognise movements from vision data. Of these, a kernel form of the SVM algorithm was most common in the studies (n = 10) (Couceiro et al, 2013;Horton et al, 2014;Kasiri-Bidhendi et al, 2015;Kasiri et al, 2017;Li et al, 2018;Montoliu et al, 2015;M. A. O'Reilly, Whelan, Ward, Delahunt, & Caulfield, 2017b;Ó Conaire et al, 2010;Reily et al, 2017;Shah et al, 2007;Zhu et al, 2006).…”

Section: Vision Recognition Model Development Methodsmentioning

confidence: 99%

“…Modality was predominately red, green, blue (RGB) cameras. Depth cameras were utilised (Kasiri-Bidhendi, Fookes, Morgan, Martin, & Sridharan, 2015;Kasiri, Fookes, Sridharan, & Morgan, 2017;Reily et al, 2017), which add depth perception for 3-dimensional image mapping.…”

Section: Vision Capture Specificationmentioning

confidence: 99%

“…Seven studies clearly reported the use of a single camera set-up (Couceiro, Dias, Mendes, & Araújo, 2013;Díaz-Pereira, Gómez-Conde, Escalona, & Olivieri, 2014;Hachaj, Ogiela, & Koptyra, 2015;Kasiri-Bidhendi et al, 2015;Kasiri et al, 2017;Nibali et al, 2017;Reily et al, 2017). One study reported 16 stationary positioned cameras at a 'bird's eye view' (Montoliu et al, 2015), andÓ Conaire et al (2010) reported the use of one overhead and 8 stationary cameras around a tennis court baseline, although data from two cameras were only used in final analysis due to occlusion issues.…”

Section: Vision Capture Specificationmentioning

confidence: 99%

“…One study reported 16 stationary positioned cameras at a 'bird's eye view' (Montoliu et al, 2015), andÓ Conaire et al (2010) reported the use of one overhead and 8 stationary cameras around a tennis court baseline, although data from two cameras were only used in final analysis due to occlusion issues. Sample frequency and, or pixel resolution were reported in seven of the studies (Couceiro et al, 2013;Hachaj et al, 2015;Kasiri-Bidhendi et al, 2015;Kasiri et al, 2017;Montoliu et al, 2015;Victor et al, 2017;Zhu et al, 2006), with sample frequencies ranging from 30 Hz to 210 Hz. Buckley et al, 2017;Buthe et al, 2016;Groh et al, 2016Groh et al, , 2017Groh et al, , 2015Jensen et al, 2016;Kautz et al, 2017;Kelly et al, 2012;Ó Conaire et al, 2010;Pernek et al, 2015;Salman et al, 2017;Schuldhaus et al, 2015;Whiteside et al, 2017).…”

Section: Vision Capture Specificationmentioning

confidence: 99%

“…Focal (Kasiri, Fookes, Sridharan, & Morgan, 2017) Swisse-range SR4000 time-of-flight (MESA Imaging AG, Switzerland)…”

mentioning

confidence: 99%

See 4 more Smart Citations

Machine and deep learning for sport-specific movement recognition: a systematic review of model development and performance

Cust

Sweeting

Ball

et al. 2018

Journal of Sports Sciences

216

148

View full text Add to dashboard Cite

Objective assessment of an athlete's performance is of importance in elite sports to facilitate detailed analysis. The implementation of automated detection and recognition of sport-specific movements overcomes the limitations associated with manual performance analysis methods. The object of this study was to systematically review the literature on machine and deep learning for sport-specific movement recognition using inertial measurement unit (IMU) and, or computer vision data inputs. A search of multiple databases was undertaken. Included studies must have investigated a sportspecific movement and analysed via machine or deep learning methods for model development. A total of 52 studies met the inclusion and exclusion criteria. Data preprocessing, processing, model development and evaluation methods varied across the studies. Model development for movement recognition were predominantly undertaken using supervised classification approaches. A kernel form of the Support Vector Machine algorithm was used in 53% of IMU and 50% of vision-based studies. Twelve studies used a deep learning method as a form of Convolutional Neural Network algorithm and one study also adopted a Long Short Term Memory architecture in their model. The adaptation of experimental setup , data pre-processing, and model development methods are best considered in relation to the characteristics of the targeted sports movement(s).

show abstract

Section: Vision Recognition Model Development Methodsmentioning

confidence: 99%

Section: Vision Capture Specificationmentioning

confidence: 99%

Section: Vision Capture Specificationmentioning

confidence: 99%

Section: Vision Capture Specificationmentioning

confidence: 99%

“…Focal (Kasiri, Fookes, Sridharan, & Morgan, 2017) Swisse-range SR4000 time-of-flight (MESA Imaging AG, Switzerland)…”

mentioning

confidence: 99%

See 3 more Smart Citations

Machine and deep learning for sport-specific movement recognition: a systematic review of model development and performance

Cust

Sweeting

Ball

et al. 2018

Journal of Sports Sciences

216

148

View full text Add to dashboard Cite

show abstract

Automatic Classification of Strike Techniques Using Limb Trajectory Data

Soekarjo

Orth²,

Warmerdam

et al. 2019

Machine Learning and Data Mining for Sports Analytics

View full text Add to dashboard Cite

General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal ? Take down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

show abstract

Human Action Recognition for Boxing Training Simulator

Anton

Makarov

2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Computer vision technologies are widely used in sports to control the quality of training. However, there are only a few approaches to recognizing the punches of a person engaged in boxing training. All existing approaches have used manual feature selection and trained on insufficient datasets. We introduce a new approach for recognizing actions in an untrimmed video based on three stages: removing frames without actions, action localization and action classification. Furthermore, we collected a sufficient dataset that contains five classes in total represented by more than 1000 punches in total. On each stage, we compared existing approaches and found the optimal model that allowed us to recognize actions in untrimmed videos with an accuracy 87%.

show abstract

Fine-grained action recognition of boxing punches from depth imagery

Cited by 25 publications

References 37 publications

Machine and deep learning for sport-specific movement recognition: a systematic review of model development and performance

Machine and deep learning for sport-specific movement recognition: a systematic review of model development and performance

Automatic Classification of Strike Techniques Using Limb Trajectory Data

Human Action Recognition for Boxing Training Simulator

Contact Info

Product

Resources

About