Dance Tempo Estimation Using a Single Leg-Attached 3D Accelerometer.

Area of Science:

• Biomechanics
• Signal Processing
• Dance Science

Background:

• Accurate dance tempo estimation is crucial for performance analysis and learning.
• Existing methods may lack precision or require multiple sensors.
• Wearable inertial devices offer a promising avenue for objective movement analysis.

Purpose of the Study:

• To develop and validate a methodology for dance tempo estimation using 3D accelerometer signals from a single wearable device.
• To assess the accuracy and feasibility of the proposed method across various solo jazz dance moves and tempos.
• To explore the potential for real-time feedback and performance assessment in dance.

Main Methods:

• Acquisition of 3D accelerometer signals from a single wearable inertial device placed on the dancer's leg.
• Implementation of an enhanced multiple resonators tempo estimation method using comb feedback filters.
• Validation using 15 solo jazz dance moves performed by professional and recreational dancers at tempos ranging from 80 to 220 bpm.

Main Results:

• The comb filter bank method achieved highly accurate dance tempo estimates, with differences of at most 1 bpm compared to dictated tempos.
• Accurate tempo estimation was demonstrated across diverse solo jazz moves, rhythmical variations, and tempos.
• The methodology proved effective even for shorter dance excerpts (e.g., four moves), enabling potential for real-time feedback.

Conclusions:

• The presented methodology accurately estimates dance tempo using a single leg-worn inertial device.
• This approach supports tempo quality and consistency assessment, aiding dance learning and performance evaluation.
• The method is extendable to other dance styles and cyclical sporting motions.