Reading Thoughts with JavaScript
summary
Live demo shows a browser page scrolled by thought using a $500 EEG module (OpenBCI), raw signals preprocessed in Python, fed into a Node.js backend and a brain.js neural network (three hidden layers in the demo). Training required ~20 hours to collect 4–5 clean samples and the model is person-specific, not generalizable out of the box. The talk walks through mapping alpha/beta/theta/delta features, uses time series filtering and FFT options, and flags strong privacy and ethics concerns for brain data.
AI-written summary · about
chapters (10)
- 0:00Start
- 1:03Overview
- 2:05Background on neurology
- 4:13Background on EEG signals
- 7:54Background on machine learning
- 10:07Mapping EEG signals
- 12:24Combining EEG signals and machine learning
- 22:42Demo
- 23:35Practical Uses
- 25:17Ethics Are Important
related talks
53mBrain: The ultimate hack
29mconsole.log(brain)
41mNeuroJavaScript
41mJavaScript lets you type with brain signals
45mMind control in JavaScript - Experimenting with alternative interactions
35mTranslating Brain Electrical Signals into Actions through Models...
46mControlling apps with your mind
30mLook mum, no hands ! Mind control in JavaScript