NCT07384065

Real-Time Avatar Feedback in Virtual Reality Exercise

Effects of Real-Time Avatar Feedback on Exercise Well-Being and Accuracy: Randomized Crossover Trial in Virtual Reality

• Status: Completed
• Phase: N/A
• Study type: Interventional
• Enrollment: 28 (actual)
• Sponsor: National Information Processing Institute · Other Government
• Sex: All
• Age: 18 Years – 50 Years
• Healthy volunteers: Accepted

Summary

The goal of this study is to examine whether real-time feedback provided by a virtual trainer avatar during virtual reality (VR) exercise influences participants' exercise well-being and movement accuracy. The study also aims to compare different types of feedback delivered by the avatar. The main questions this study aims to answer are: Does real-time avatar feedback improve exercise well-being during VR-based training? Does real-time avatar feedback improve movement accuracy during exercise? Are there differences between no feedback, verbal feedback, and combined verbal-gestural feedback? Researchers will compare three VR exercise conditions: no feedback, verbal feedback, and combined verbal-gestural feedback, using a randomized crossover design so that each participant experiences all conditions. Participants will: Take part in three VR exercise sessions, each lasting approximately 15 minutes Complete the sessions under different feedback conditions, with the order randomized Perform guided physical exercises while wearing a VR headset and following a virtual trainer avatar Provide self-reported ratings of exercise experience after each session This study is an interventional, randomized crossover trial conducted in adults, with all procedures completed during a single study period.

Detailed description

his study investigates the effects of real-time feedback delivered by a virtual trainer avatar during exercise performed in a virtual reality (VR) environment. The intervention is designed to examine how different modes of avatar feedback influence the exercise experience and movement performance during short, guided training sessions. The study uses a randomized crossover design, in which each participant completes multiple VR exercise sessions under different feedback conditions. The order of conditions is randomized to control for sequence and learning effects. The feedback conditions differ only in the type of information provided by the avatar and include: no feedback, verbal feedback, and combined verbal-gestural feedback. All other aspects of the VR environment, exercise content, and session duration are kept constant across conditions. The VR application presents a virtual trainer avatar that demonstrates and guides a standardized set of physical exercises. A custom algorithm processes real-time motion data captured by the VR system, including parameters related to movement pace and range of motion. Based on this analysis, the avatar delivers adaptive feedback during the exercise session according to the assigned condition. In the no-feedback condition, the avatar performs the exercises without providing corrective or motivational input. Each VR session lasts approximately 15 minutes and is separated from the subsequent session by a fixed 7-day interval to minimize carryover effects. The study is conducted within a single study period, and all procedures are completed after participants finish the final VR session. The intervention does not involve any pharmacological treatment or medical devices regulated for clinical use; VR hardware is used solely as a research tool to deliver the exercise environment and capture movement data. This study is classified as an interventional behavioral trial and is intended to contribute to research on virtual trainers, human-computer interaction, and technology-supported physical activity by systematically examining the role of real-time avatar feedback in VR-based exercise.

Conditions

• Healthy Participants

Interventions

Type	Name	Description
BEHAVIORAL	VR Exercise With No Avatar Feedback	Participants perform a standardized exercise session in a virtual reality environment guided by a virtual trainer avatar. The avatar demonstrates the exercises but does not provide any real-time verbal or gestural feedback during the session.
BEHAVIORAL	VR Exercise With Verbal Avatar Feedback	Participants perform a standardized exercise session in a virtual reality environment guided by a virtual trainer avatar. Real-time motion data are analyzed by the VR system, and the avatar provides adaptive verbal feedback during the exercise session.
BEHAVIORAL	VR Exercise With Verbal and Gestural Avatar Feedback	Participants perform a standardized exercise session in a virtual reality environment guided by a virtual trainer avatar. Real-time motion data are analyzed by the VR system, and the avatar provides adaptive verbal and gestural feedback during the exercise session.

Timeline

• Start date: 2024-05-06
• Primary completion: 2024-05-24
• Completion: 2024-05-24
• First posted: 2026-02-03
• Last updated: 2026-02-03

Locations

1 site across 1 country: Poland