Yoga Posture Detection and Correction Using YOLOv8

Abstract: This paper presents a novel approach to real- time yoga posture detection and correction using the YOLOv8 objectdetectionalgorithm.Thesystemaimsto assist practitioners in performing yoga poses correctly, thereby reducing the risk of injuries and enhancing the effectiveness of their practice. The system works by analyzinglivevideo feedsofpractitionersandcomparing theirposestoadatabaseofcorrectposes.Ifdiscrepancies are detected, real-time feedbackis provided to guide the practitioner in adjusting their posture. The system also tracksthepractitioner's progressovertime, allowing for personalized feedback and recommendations. Experimentalresultsdemonstratetheeffectivenessofthe system in accurately detecting and correcting yoga postures, highlighting its potential to revolutionize the way yoga is practiced and taught