Kyudo Practice System Using VR and Motion Sensors

This paper presents a Virtual Reality (VR)-based kyudo form practice system using wearable inertial measurement unit (IMU) sensors and hand tracking to support solo training of Syaho-Hassetsu (the eight shooting stages). The system estimates the practitioner’s motion, segments it into stages with a time-aware finite-state machine, and provides stage-specific visual guidance inside a head-mounted display. To reduce individual differences, the system introduces a simple user measurement (fingertip-to-body midline length) to scale key distances such as stance width and draw length thresholds. The implementation is built in Unity and designed for real-time use on a standalone VR headset connected to an external motion-sensing pipeline. We position this work as a system-implementation and feasibility study: rather than asserting training effectiveness, our aim is to demonstrate that the proposed system can be realized and to report initial feasibility observations. In a small preliminary evaluation, results suggest VR guidance may improve stance-width consistency, and we treat these findings as feasibility evidence rather than conclusive proof of effectiveness.