SPD domain-specific batch normalization to crack interpretable unsupervised domain adaptation in EEG
SPD domain-specific batch normalization to crack interpretable unsupervised domain adaptation in EEG
复制标题
DOI：
10.48550/arxiv.2206.01323
复制
发表时间：
2022-06
期刊：
影响因子：
--
通讯作者：
Reinmar J. Kobler;J. Hirayama;Qibin Zhao;M. Kawanabe
中科院分区：
文献类型：
--
作者：
Reinmar J. Kobler;J. Hirayama;Qibin Zhao;M. Kawanabe
脑电图（EEG）以毫秒级的分辨率非侵入性地提供对神经元动力学的访问，使其成为神经科学和医疗保健中的可行方法。然而，其效用是有限的，因为当前EEG技术不能很好地跨域推广（即，会话和主题）而无需昂贵的监督重新校准。当代方法将这种迁移学习（TL）问题视为多源/目标无监督域自适应（UDA）问题，并使用深度学习或浅层黎曼几何感知对齐方法来解决它。到目前为止，这两个方向都未能始终关闭性能差距，以最先进的领域特定的方法的基础上切空间映射（TSM）的对称正定（SPD）流形。在这里，我们提出了一个基于理论的机器学习框架，它首次以端到端的方式学习领域不变的TSM模型。为了实现这一目标，我们提出了一种新的几何深度学习构建块，我们将其表示为SPD特定域动量批量归一化（SPDDSMBN）。SPDDSMBN层可以将特定于域的SPD输入转换为域不变的SPD输出，并且可以容易地应用于多源/目标和在线UDA场景。在6个不同的EEG脑机接口（BCI）数据集的广泛实验中，我们获得了最先进的性能在会话间和主题TL与一个简单的，本质上可解释的网络架构，我们表示TSMNet。
Electroencephalography (EEG) provides access to neuronal dynamics non-invasively with millisecond resolution, rendering it a viable method in neuroscience and healthcare. However, its utility is limited as current EEG technology does not generalize well across domains (i.e., sessions and subjects) without expensive supervised re-calibration. Contemporary methods cast this transfer learning (TL) problem as a multi-source/-target unsupervised domain adaptation (UDA) problem and address it with deep learning or shallow, Riemannian geometry aware alignment methods. Both directions have, so far, failed to consistently close the performance gap to state-of-the-art domain-specific methods based on tangent space mapping (TSM) on the symmetric positive definite (SPD) manifold. Here, we propose a theory-based machine learning framework that enables, for the first time, learning domain-invariant TSM models in an end-to-end fashion. To achieve this, we propose a new building block for geometric deep learning, which we denote SPD domain-specific momentum batch normalization (SPDDSMBN). A SPDDSMBN layer can transform domain-specific SPD inputs into domain-invariant SPD outputs, and can be readily applied to multi-source/-target and online UDA scenarios. In extensive experiments with 6 diverse EEG brain-computer interface (BCI) datasets, we obtain state-of-the-art performance in inter-session and -subject TL with a simple, intrinsically interpretable network architecture, which we denote TSMNet.