Filter Bank Common Spatial Pattern (FBCSP) in Brain-Computer Interface

By Kai Keng Ang, Z. Y. Chin, Haihong Zhang, Cuntai Guan. IEEE World Congress on Computational Intelligence, 2008

1,466 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/ab6a90f44f0fe280f6207981bb62beb88602281e

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Filter Bank Common Spatial Pattern Algorithm on BCI Competition IV Datasets 2a and 2b (2012, methodology): “Given that ω , ω ∈ {1, 2, 3, 4} represents the left, right, foot, and tongue motor imagery, the Divide-and-Conquer (DC) approach adopts a tree-based classiﬁer approach (Zhang et al., 2007; Chin et al.,2009).”
EEG-Based Spatio–Temporal Convolutional Neural Network for Driver Fatigue Evaluation (2019, unknown): “The filter bank CSP method in [21] was extended to help improve the performance of signal decomposition.”
Review of the BCI Competition IV (2012, methodology): “is based on the filter bank common spatial pattern (FBCSP) variant (Ang et al., 2008).”
Learning Temporal Information for Brain-Computer Interface Using Convolutional Neural Networks (2018, methodology): “FBCSP was ﬁrst introduced as an extension to the original Common Spatial Patterns (CSP) algorithm and gained attention by winning the 2008 BCI Competition IV-2a [16], [15].”
EEG-TCNet: An Accurate Temporal Convolutional Network for Embedded Motor-Imagery Brain–Machine Interfaces (2020, methodology): “One of the ﬁrst successful CNN in MI classiﬁcation was Shallow ConvNet [9], which was inspired by FBCSP.”
Subject-Independent Brain–Computer Interfaces Based on Deep Convolutional Neural Networks (2019, methodology): “The top row depicts the subject-dependent approaches in the order of CSP, FBCSP, and BSSFO, and the bottom row displays the 6 and 7 show the performance changes by the number of subjects and the frequency bands.”
EEG dataset and OpenBMI toolbox for three BCI paradigms: an investigation into BCI illiteracy (2019, unknown): “To date, a whole range of methods have been proposed that fulfill this task, such as heuristics [54], filter-banks [25,55], Bayesian methods [26], among many others.”
Mental workload during n-back task—quantified in the prefrontal cortex using fNIRS (2014, methodology): “See Ang et al. (2008) for a more detailed description of feature selection methods using Mutual Information.”
Optimizing the Channel Selection and Classification Accuracy in EEG-Based BCI (2011, methodology): “Comparing Tables III and IV also shows that on average, dataset IVa required more channels to achieve higher accuracy compared to dataset IIa.”
Regularized Common Spatial Pattern With Aggregation for EEG Classification in Small-Sample Setting (2010, methodology): “In [35], EEG signals are decomposed into subbands, where CSP is applied to extract features from each subband, and then the subband scores are”
HS-CNN: a CNN with hybrid convolution scale for EEG motor imagery classification (2020, background): “To address this issue, the authors of [35] proposed filter bank common spatial pattern (FBCSP), an improved version of CSP.”
EEGLAB, SIFT, NFT, BCILAB, and ERICA: New Tools for Advanced EEG Processing (2011, methodology): “…Regularized and variational Bayesian logistic regression and sparse Bayesian logistic regression [39, 40] (vi) Baseline ﬁltering (4) Multiband-CSP [41] (1) Hierarchical kernel learning [42] (vii) Resampling (5) Multi-Model Independent component features (viii) Relevance vector machines (RVM) [43]…”
BCILAB: a platform for brain–computer interface development (2013, background): “While these CSP variants operate on a single frequency band (usually a ‘wideband’ 7-30 Hz window for motor imagery task data), the implemented variant of the Filter-Bank CSP approach (FBCSP) [57] instead solves one CSP instance per frequency band of interest and pools the resulting features.”
Transfer Learning for EEG-Based Brain–Computer Interfaces: A Review of Progress Made Since 2016 (2020, unknown): “For example, Schirrmeister et al. [65] proposed two convolutional neural networks (CNNs) for EEG decoding, and showed that both outperformed ﬁlter bank common spatial patterns (FBCSP) [66] in cross-subject MI classiﬁcation.”
An end-to-end deep learning approach to MI-EEG signal classification for BCIs (2018, background): “On the one hand, this suggests that low-pass ﬁltering the signal before processing, as it is often done (refer to (Ang et al., 2008; Bentlemsan et al., 2014; Park et al., 2014a; Tabar & Halici, 2017)), may even discard relevant information and degrade the classiﬁcation results.”
A New Discriminative Common Spatial Pattern Method for Motor Imagery Brain–Computer Interfaces (2009, methodology): “The FBCSP in [6] algorithm extracts CSP features from a fixed FB consisting of nine Chebyshev type II bandpass filters and a feature selection process is done before classifying the signals.”
Macroscopic EEG reveals discriminative low-frequency oscillations in plan-to-grasp visuomotor tasks (2025, unknown): “…Grasp vs no grasp scenarios: • Pen (60 trials) vs empty (60 trials) • Bottle (60 trials) vs empty (60 trials) For multiclass classification, CSP filters were extended to handle all three conditions simultaneously using the one-vs-rest strategy [37] given the binary constraint of the FBCSP method.”