← All research projects

Muse Flow: A Web-Based Humanistic Intelligence System for Audio Mediated Reality

27th Annual Mersivity / WaterHCI Symposium, December 2025
Toronto and online, December 4 to 21, 2025 · Published in the symposium proceedings (open access on Zenodo)

Four students presenting on a stage beside a transparent ball, a slide showing a relaxation index behind them

The Muse Flow team presenting at the December 2025 symposium, with the relaxation index on the screen and a Mersivity ball on stage.

Muse Flow closes a loop between your brain and your ears. A Muse 2 headband streams EEG to a web page, which computes a relaxation index from alpha and beta power. When you are stressed or hard at work the music fades in; when you calm down it fades out. It is a small, transparent example of Mann's Humanistic Intelligence: the computer and the person in one feedback loop, nothing hidden.

Abstract

Muse Flow is a web-based system implementing Humanistic Intelligence (HI) through a continuous feedback loop between a user's cognitive state and their auditory environment. Using the Muse EEG headband, the system dynamically modulates music playback based on real-time neural activity, creating an Audio Mediated Reality that augments both focus and relaxation. The system computes a Relaxation Index based on the ratio of Alpha to Alpha+Beta activity. When the user becomes stressed or highly focused (low Relaxation Index), music fades in; when the user becomes calm (high Relaxation Index), music fades out. The project contributes (i) a lightweight client-side EEG pipeline for consumer-grade headsets, (ii) a transparent focus/relaxation control policy suitable for real-time neurofeedback, and (iii) a human-centered design rationale grounded in HI and contemporary neurotechnology ethics; in particular, the team aims to uphold Mersivity, which is the building of technology to unite individuals with each other and with nature. All processing is done locally in the browser, emphasizing privacy and user agency. Code and demo available at: https://github.com/GIOGIO156/ECE1724.git

Authors

Hao-Chih Huang, Yiwei Cao, Jinghan Song, Huiying Yu, Alexander Vicol, Steve Mann

Links

← All research projects