DS-MTNet: Structured multi-task EEG decoding for human-machine collaboration

doi:10.1016/j.neucom.2026.135222

DS-MTNet: Structured multi-task EEG decoding for human-machine collaboration

2026-09-23

0

PRE

AI

DOI：10.1016/j.neucom.2026.135222

Original

Original request for help

Save

Abstract

En 中文

• This study conceptualized the human operator as an active information source to complement machine-side sensing in HMC. • DS-MTNet selectively routed fine-grained evidence, reorganizing and reusing complementary information retained across stages. • DS-MTNet achieved the highest mean performance on the main dataset; ablation and routing analyses, together with validation on three additional public datasets, supported its structured, task-selective representation design.

Keywords:

Electroencephalography (EEG)

Human-machine collaboration

EEG decoding

Structured representation learning

Selective routing

DS-MTNet: Structured multi-task EEG decoding for human-machine collaboration

Abstract

Journal

Organization

Cited Papers