{"title":"Personal Exercise Assistant: Correcting Exercise Posture using Modified Open pose","authors":"Manisha C. Patel, Nilesh B. Kalani","doi":"10.17762/msea.v71i4.1871","DOIUrl":null,"url":null,"abstract":"Fitness exercises or yoga are very bene?cial to personal health and ?tness; however, they can also be ineffective and potentially dangerous if performed in correctly by the user. Exercise mistakes are made when the user does not use the proper form, or pose. In our work, we introduce Personal Exercise Assistant, an application that detects the user’s exercise pose and provides personalized feedback. Modified Open Pose is used as a Pose estimator in this application. Open Pose is a pre trained model composed of a multi-stage CNN to detect a user’s posture. This application then evaluates the vector geometry of the pose through an exercise to provide helpful feedback. Pose estimation is a method in which spatial locations of key body joints is calculated using image or video of the person. Modified open pose is a light weight (8bit Quantized version) Open pose. Personal Exercise Assistant works on yoga exercises and supports any Windows or Linux computer with a GPU.","PeriodicalId":37943,"journal":{"name":"Philippine Statistician","volume":" ","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2022-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Philippine Statistician","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.17762/msea.v71i4.1871","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"Mathematics","Score":null,"Total":0}
引用次数: 1
Abstract
Fitness exercises or yoga are very bene?cial to personal health and ?tness; however, they can also be ineffective and potentially dangerous if performed in correctly by the user. Exercise mistakes are made when the user does not use the proper form, or pose. In our work, we introduce Personal Exercise Assistant, an application that detects the user’s exercise pose and provides personalized feedback. Modified Open Pose is used as a Pose estimator in this application. Open Pose is a pre trained model composed of a multi-stage CNN to detect a user’s posture. This application then evaluates the vector geometry of the pose through an exercise to provide helpful feedback. Pose estimation is a method in which spatial locations of key body joints is calculated using image or video of the person. Modified open pose is a light weight (8bit Quantized version) Open pose. Personal Exercise Assistant works on yoga exercises and supports any Windows or Linux computer with a GPU.
期刊介绍:
The Journal aims to provide a media for the dissemination of research by statisticians and researchers using statistical method in resolving their research problems. While a broad spectrum of topics will be entertained, those with original contribution to the statistical science or those that illustrates novel applications of statistics in solving real-life problems will be prioritized. The scope includes, but is not limited to the following topics: Official Statistics Computational Statistics Simulation Studies Mathematical Statistics Survey Sampling Statistics Education Time Series Analysis Biostatistics Nonparametric Methods Experimental Designs and Analysis Econometric Theory and Applications Other Applications