MPNet: A Robust and Efficient Manifold Pooling Network for Multi-Rhythm EEG Signal Decoding

TL;DR

MPNet is a novel manifold pooling network that efficiently decodes multi-rhythm EEG signals by combining a rhythm-adaptive frontend with a pooling layer, achieving state-of-the-art accuracy and faster processing.

Contribution

The paper introduces MPNet, a new deep Riemannian network with a manifold node pooling layer that reduces computational costs while maintaining high decoding accuracy.

Findings

MPNet achieves state-of-the-art accuracy on public EEG datasets.

MPNet runs up to 10 times faster than comparable Riemannian models.

MPNet maintains robust performance under limited-data conditions.

Abstract

Deep Riemannian networks provide a powerful framework for Electroencephalography (EEG) decoding, but their practical applications are severely constrained. Accurately decoding EEG signals requires modeling complex temporal dynamics across multiple rhythms, which results in high-dimensional Riemannian inputs and significant computational costs. To address this, we propose the Manifold Pooling Network (MPNet). MPNet uses a rhythm-adaptive convolutional frontend to extract comprehensive time-frequency representations and generate multi-view Riemannian nodes. A novel manifold node pooling layer is then proposed to aggregate these nodes into a single fusion node with a fixed size, enabling the following deep Riemannian network to process it with greatly reduced costs. Experiments on two public EEG datasets show that MPNet achieves state-of-the-art accuracy, runs up to 10 times faster than the comparable Riemannian model, and maintains robust performance under limited-data conditions. These findings highlight MPNet's practicality and efficiency for real-world EEG applications.