am-mteeg-classification

What problem does it solve?

EEG data from multiple individuals often exhibit significant inter-subject variability, making cross-subject classification challenging; AM-MTEEG provides a hippocampus-inspired, associative-memory based multi-task framework to harmonize these differences and improve generalization.

Core Features & Use Cases

• Cross-subject BCI classification with reduced inter-subject variability
• Multitask adapters for subject-specific nuances while sharing shared features
• Interpretable waveform representations through spiking-like encoding and memory associations
• End-to-end EEG pipeline: encoding, spiking representation, and associative memory classification

Quick Start

Run AM-MTEEG on your EEG dataset to perform cross-subject classification using associative memory with interpretable waveform representations