Recognizing Solo Jazz Dance Moves Using a Single Leg-Attached Inertial Wearable Device.

Area of Science:

• Biomechanics
• Wearable Technology
• Signal Processing

Background:

• Automated recognition of human movement, particularly complex activities like dance, remains a challenge.
• Existing methods often struggle with variations in tempo, orientation, and performer skill.

Purpose of the Study:

• To develop and validate a robust method for recognizing dance moves from wearable sensor data.
• To create a system that is independent of dance tempo and device orientation for broad applicability.

Main Methods:

• Utilized 3D accelerometer and gyroscope data from a single leg-worn device.
• Implemented dance tempo estimation, temporal scaling, and orientation-invariant coordinate transformation.
• Employed sliding correlation-based template matching for move recognition.

Main Results:

• Achieved high recognition F1 scores (0.87–0.98) across various tempos and dancers.
• Demonstrated accuracy in recognizing repeated solo jazz dance moves.
• Validated the method's potential for recognizing imitated dance sequences.

Conclusions:

• The presented method offers a reliable approach for automated dance move recognition.
• The system's independence from tempo and orientation enhances its practical usability.
• This technology can support dance performance analysis and skill acquisition.