Sounds Aware: A Mobile App for Raising Awareness of Environmental Sound
summary
A web app runs on a phone, extracts MFCC features with Meyda on-device, classifies sounds with a simple k‑nearest neighbor model and maps detected anthrophony to ambient music while fading it when biophony or geophony rises. Users can add labeled examples for personalization; playback uses p5.js/tone and requires headphones so microphone doesn’t pick up output. A small study and developer use report show increased attention to nonhuman sounds and a behavioral change where the app’s builder eventually stopped needing headphones.
AI-written summary · about
chapters (5)
- 0:00Start
- 0:54How I Got Started Thinking about this App
- 6:38Ubiquitous Listen
- 7:36Audio Walk
- 7:54The Quiet Walk
related talks
01h 14mExploring the hidden potential of sound data
20mExploring the hidden potential of sound data
17mTechnology: A Changemaker in Environmental Protection
06mSensing to Hear: Speech Enhancement for Mobile Devices Using Acoustic Signals
29mBioacoustics and Machine Learning as Key Tools in Conservation
25mThe Sound of Life
40mWho's singing? Automatic bird sound recognition with machine learning
26mMaking Sense The Road to mobile Awareness