Excellent fine-tuning: From specific-subject classification to cross-task classification for motor imagery. (January 2023)

• Record Type: Journal Article
• Authors: Jia, Xueyu; Song, Yonghao; Xie, Longhan
• Is Part Of: Biomedical signal processing and control. Volume 79(2023)Part 1
• Publication Date: 2023-01
• Subjects: Motor imagery -- Common spatial pattern -- Transformer -- Metric learning -- Cross subjects; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2022.104051
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: Full-scale tests on continuous steel-UHPC composite slabs (SUCS) were conducted. Elaborate numerical model of SUCS was built and validated. A new constitutive model was proposed to account for the strain-hardening UHPC. Comprehensive parametric studies were conducted using the validated numerical model. Abstract: With the popularity of deep learning, motor imagery electroencephalogram (MI-EEG) recognition based on feature extractors and classifiers has performed well. However, the features extracted by most models are not discriminative enough and are limited to specific-subject classifi-cation. We proposed a novel model Metric-based Spatial Filtering Transformer (MSFT) that utilizes additive angular margin loss to enforce the deep model to improve inter-class separability while enhancing intra-class compactness. Besides, a data augmentation method called EEG pyramid was applied to the model. Our model not only outperforms many recent benchmarks in specific-subject classifi-cation, but also is used for cross-subject and even cross-task classification. We did some experiments using BCI competition IV 2a and 2b datasets to evaluate the average accuracy. The Specific-subject : 86.11 % for 2a, 88.39 % for 2b. The Cross-subject : 61.92 % for 2a. The Cross-task : training the feature extractor with 2a data and then fine-tuning the classifier with 2b can achieve an average accuracy of 83.38 %. Our method is more general than most benchmarks and can deal with different kinds of classification situations.