# Optimizing the Interaction System for Treadmill Video Games Using a Smartphone’s Front Camera

## TL;DR

This paper presents a new system for treadmill video games that uses a smartphone's front camera to track facial movements for hands-free control.

## Contribution

The paper introduces a novel FFT-based method for real-time cadence estimation using only a smartphone's front camera.

## Key findings

- The FFT-based cadence estimation achieves less than 5% error using only 128 frames.
- The system outperforms traditional peak detection in accuracy and robustness across various running speeds.
- The solution is practical, efficient, and scalable for fitness-oriented human–computer interaction.

## Abstract

This paper introduces a lightweight and accessible interaction system for treadmill-based video games, relying solely on facial tracking via a smartphone’s front camera. The system enables real-time estimation of running cadence and directional control through natural head movements, providing an immersive and hands-free gaming experience. A key contribution is the implementation of a FFT-based cadence estimation method that achieves accuracy errors below 5% using only 128 frames, enabling real-time feedback. Preliminary evaluations with 11 participants demonstrate that the FFT-based approach outperforms traditional peak detection in both accuracy and robustness across multiple running speeds. These results position the system as a practical, efficient, and scalable solution for fitness-oriented human–computer interaction, with promising implications for digital health and exergaming.