Multi-Domain Feature Fusion and Channel Attention in an Inception-Based Architecture for Motor Imagery EEG Decoding
The results reveal that the proposed framework can distinguish motor imagery patterns across different limb regions and effectively classify predictive EEG features between symmetric limbs performing identical imagined movements, providing an offline proof-of-concept for multi-limb BCI control.