发表机构
Université Paris-Saclay; Yneuro; University of California San Diego; Imperial College London（巴黎萨克雷大学; Yneuro公司; 加州大学圣地亚哥分校; 帝国理工学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究在EEG时间序列解码中评估神经元级架构增长，发现增长在能有效排序候选神经元时提升性能，如ShallowFBCSPNet以半参数获得2.9点提升，并指出跳过添加率可指导小规模从头训练。
AI 中文摘要
卷积EEG解码器以固定宽度进行训练，该宽度通常由其作者在其他数据上设定。增长方法在训练过程中添加神经元，在损失可能下降最多的位置进行添加，但相较于参考宽度，这些方法是否有所改进在EEG上尚未得到验证。在此，我们在12个运动想象数据集上，在三种协议下增长三个卷积骨干网络，并将每个网络与其每个受试者的参考模型进行比较。增长的ShallowFBCSPNet得分比其参考模型高出2.9个百分点，而参数仅为后者的一半（0.57倍）；SCCNet的变化最多为1.2个百分点。Deep4Net增长模型显示出准确率下降，但它们需要适应性调整，这妨碍了结果的忠实比较。这些差异遵循选择步骤，该步骤通过奇异值分解的动态阈值来保留候选神经元。总体而言，这些结果表明，当增长标准能够对候选神经元进行排序时，增长是有帮助的，并且跳过神经元添加的比率可以指示解码器在何处可以从零开始小规模增长。
英文摘要
Convolutional EEG decoders are trained at a fixed width, usually set by their authors on other data. Growing methods add neurons during training where the loss could decrease the most, but whether they improve compared to a reference width is untested on EEG. Here, we grow three convolutional backbones on 12 motor-imagery datasets under three protocols and compare each with its reference model per subject. The growing ShallowFBCSPNet scores 2.9 points above its reference model with only half the parameters (0.57x), SCCNet changes by at most 1.2 points. Deep4Net growing models show decreased accuracy, but they require adaptation that prevent to compare faithfully the results. These differences follow the selection step, which keeps a candidate neuron relying on a dynamic threshold from singular values decomposition. Overall, these results suggest that growth helps when its criterion can rank the candidate neurons, and that the rate of skipped neuron addition tells where a decoder can be grown small from scratch.
Comments5 pages, 3 figures, 2 tables. Submitted to ICASSP 2027
