DSAINet: An Efficient Dual-Scale Attentive Interaction Network for General EEG Decoding

2026-04-20

AI Summary

Existing EEG decoding models exhibit limited generalization in cross-task and subject-independent settings due to task-specific temporal inductive biases. To address this, this work proposes DSAINet—a lightweight, general-purpose EEG decoding architecture that leverages shared spatiotemporal token representations and employs parallel fine-grained and coarse-grained convolutional branches to capture multiscale temporal dynamics. The model further integrates intra- and inter-branch adaptive attention mechanisms to effectively fuse task-relevant features. Notably, DSAINet achieves multi-task EEG decoding within a unified framework without task-specific customization for the first time. Evaluated across ten public datasets and five downstream tasks, it consistently outperforms thirteen baseline models, achieving state-of-the-art accuracy-efficiency trade-offs with only approximately 77,000 parameters, while also offering interpretable neurophysiological insights.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Deep Neural Architectures and Foundation Models
• Humans and AI: Brain-Sensing and Analysis

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems

Abstract

In real-world applications of noninvasive electroencephalography (EEG), specialized decoders often show limited generalizability across diverse tasks under subject-independent settings. One central challenge is that task-relevant EEG signals often follow different temporal organization patterns across tasks, while many existing methods rely on task-tailored architectural designs that introduce task-specific temporal inductive biases. This mismatch makes it difficult to adapt temporal modeling across tasks without changing the model configuration. To address these challenges, we propose DSAINet, an efficient dual-scale attentive interaction network for general EEG decoding. Specifically, DSAINet constructs shared spatiotemporal token representations from raw EEG signals and models diverse temporal dynamics through parallel convolutional branches at fine and coarse scales. The resulting representations are then adaptively refined by intra-branch attention to emphasize salient scale-specific patterns and by inter-branch attention to integrate task-relevant features across scales, followed by adaptive token aggregation to yield a compact representation for prediction. Extensive experiments on five downstream EEG decoding tasks across ten public datasets show that DSAINet consistently outperforms 13 representative baselines under strict subject-independent evaluation. Notably, this performance is achieved using the same architecture hyperparameters across datasets. Moreover, DSAINet achieves a favorable accuracy-efficiency trade-off with only about 77K trainable parameters and provides interpretable neurophysiological insights. The code is publicly available at https://github.com/zy0929/DSAINet.

Problem

• EEG decoding
• subject-independent
• temporal dynamics
• generalizability
• task-specific bias

Innovation

• dual-scale attention
• general EEG decoding
• subject-independent
• spatiotemporal token representation
• adaptive feature aggregation