SwitchBraidNet: Quantisation-Aware Lightweight Architecture for Hybrid Brain-Computer Interface

2026-06-17

AI Summary

This work addresses the challenge of deploying hybrid brain–computer interfaces—combining motor imagery (MI) and steady-state visual evoked potentials (SSVEP)—on low-power embedded platforms due to their high computational complexity. To this end, the authors propose SwitchBraidNet, a lightweight EEG classification architecture that integrates a dual-path temporal braiding structure to capture multi-scale oscillatory features, an adaptive spatial switch for electrode gating, and a log-variance readout layer that directly encodes bandpower. The model supports quantization-aware training across FP32, FP16, and INT8 precisions. Evaluated on the OpenBMI dataset, it achieves 69.49% MI accuracy (FP16), 93.48% SSVEP accuracy (FP32), and a hybrid information transfer rate of 64.82 bits/min (FP16), while the INT8 version occupies only 3.03 KB, substantially enhancing the feasibility of hybrid BCI deployment on resource-constrained devices.

Abstract

Hybrid brain-computer interfaces (BCIs) that integrate motor imagery (MI) and steady-state visual evoked potentials (SSVEP) provide high-dimensional neural decoding but typically exceed the computational limits of embedded hardware. To address this, we propose SwitchBraidNet, a compact EEG classification architecture designed for low-power deployment. The model employs a dual-path temporal braid to extract multiscale oscillatory features, an adaptive squeeze-and-excitation spatial switch for electrode gating, and a log-variance readout layer for direct band-power encoding. Furthermore, through systematic quantisation-aware training on the OpenBMI dataset, we compared SwitchBraidNet against four established baselines across FP32, FP16, and INT8 precisions. Experimental results demonstrate superior efficiency and performance, achieving MI accuracy of 69.49% (FP16), SSVEP accuracy of 93.48% (FP32), and a hybrid information transfer rate of 64.82 bits/min (FP16). With an INT8 footprint of only 3.03 KB, SwitchBraidNet maintains high accuracy across varying numerical precisions, demonstrating its suitability for low-power embedded BCI deployment.

Problem

• hybrid brain-computer interface
• computational constraints
• embedded hardware
• EEG classification
• low-power deployment

Innovation

• quantisation-aware training
• lightweight architecture
• hybrid BCI
• temporal braid
• spatial switch