Zhenai Wang¹ and Matthew Ngoi², ¹USA, ²California State Polytechnic University, USA

This paper addresses the challenge of subjectivity in dance performance analysis and the lack of standardized feedback [4]. We introduce a computer vision-based program that offers a systematic framework for assessing posture disparities within dance routines [5]. Through advanced motion tracking algorithms, the program objectively identifies and contrasts key postures and movements. Our experiments involving intermediate and novice dancers demonstrate significant skill improvements, accelerated learning speed, and positive user experiences. Although limitations in mirror material hindered optimal user viewing distance, future plans involve integrating technology within a physical mirror to provide unrestricted comprehensive feedback. This work offers a promising avenue for enhancing dance education by providing consistent, real-time feedback, bridging the gap between subjective assessments and standardized analysis to empower dancers in their skill development [6].

Dance, Pose estimation (Mediapipe), Body tracking, real-time