Key Results
TODO_ACCURACY_PERCENT% cross-subject accuracy (TODO_CV_PROTOCOL)
Macro F1: TODO_F1_MACRO
Subjects: TODO_N_SUBJECTS | Trials: TODO_N_TRIALS | Latency: TODO_LATENCY_MS ms
AlphaHand
EEG Finger Movement Identification Research
Individual Finger Movement Identification From Muse 2 EEG
A lightweight EEG pipeline for decoding individual finger intent with Muse 2 signals.
AlphaHand is an EEG finger movement identification effort using Muse 2 signals to decode individual finger intent for practical BCI and HCI applications.
Pipeline at a glance
Capture multi-channel Muse 2 EEG, process in real time, and decode finger intent with a lightweight neural stack.
Adaptive filteringTemporal featuresDeployment ready
Capture
Muse 2 headset + finger prompts with synchronized annotations.
Decode
Feature extraction + temporal model trained for individual intent.
Deploy
Low-latency inference ready for assistive and HCI systems.
Read the paper
Publication-ready research
The manuscript covers experimental design, preprocessing, model architecture, and evaluation results.
Highlights
- • Multi-subject evaluation with robust cross-validation.
- • Feature visualization aligned with neurophysiological markers.
- • Portable inference pipeline for real-world deployments.
Media highlights
Latest visuals & demos
Signal Capture
Muse 2 EEG acquisition with motion prompts.
Model Training
Temporal convolution + transformer classifier.