Design and Experimental Study of a Dance Imitation Scoring System Based on Computer Vision
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
This research introduces a new dance imitation scoring system leveraging computer vision techniques to assess dance performance quality. The proposed system utilizes sophisticated pose estimation methods, such as OpenPose and MediaPipe, to identify human body key points and track movements within dance videos. The system evaluates essential aspects of dance imitation, including pose accuracy, timing synchronization, and movement coordination. By combining dynamic time warping with deep learning models, the system compares imitated dance routines to reference performances, generating a quantitative score. Experimental results demonstrate that the system achieves high accuracy, with results aligning closely with human judges across various dance styles, including ballet, hip-hop, and contemporary dance. The system offers significant advantages in terms of automation, precision, and consistency when compared to traditional manual scoring methods. However, challenges remain in handling rapid movements, occlusions, and ensuring real-time performance. Proposed improvements include the integration of 3D pose estimation and multi-angle tracking. This system has wide applications across entertainment, virtual reality, and personalized dance training, presenting new possibilities for automated dance assessment.
Fields of Study & Subjects
Human Pose and Action Recognition
Human Motion and Animation
Hand Gesture Recognition Systems
Publication details
- Journal / source
- IEEE International Conference on Power, Electronics and Green Energy (ICPEGE)
- Pages
- 1104–1111
- Publication date
- 28 January 2026
- Language
- EN