Enhancing SSVEP Identification with Less Individual Calibration Data Using Periodically Repeated Component Analysis DOI Creative Commons
Yufeng Ke, Shuang Liu, Dong Ming

et al.

Published: July 20, 2023

<p> Spatial filtering and template matching-based methods are commonly used to identify the stimulus frequency from multichannel EEG signals in steady-state visually evoked potentials (SSVEP)-based brain-computer interfaces (BCIs). However, these require sufficient calibration data obtain reliable spatial filters SSVEP templates, they underperform identification with small-sample-size data, especially when a single trial of is available for each frequency. In contrast state-of-the-art task-related component analysis (TRCA)-based methods, which construct templates based on inter-trial components SSVEP, this study proposes method called periodically repeated (PRCA), constructs maximize reproducibility across periods synthetic by replicating (PRCs). We also introduced PRCs into two improved variants TRCA. Performance evaluation was conducted using self-collected 16-target dataset public 40-target dataset. The proposed show significant improvements less training can achieve comparable performance baseline 5 trials 2 or 3 trials. Using frequency, PRCA-based achieved highest average accuracies over 95% 90% 1-s length maximum information transfer rates 198.8±57.3 bits/min 191.2±48.1 sets, respectively. Our results demonstrate effectiveness robustness reduced effort suggest its potential practical applications SSVEP-BCIs. </p>

Language: Английский

Memristor chip-enabled adaptive neuromorphic decoder for co-evolutional brain-computer interfaces DOI Creative Commons
Huaqiang Wu, Zhengwu Liu, Jie Mei

et al.

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: March 4, 2024

Abstract To fulfill complex human-machine interactions, a brain-computer interface (BCI) must not only decipher brain signals but also dynamically adapt to fluctuations, ultimately co-evolving with the brain. This necessitates novel decoder capable of flexible updates energy-efficient decoding capabilities. In this work, we designed co-evolutional BCI neuromorphic enabled by 128k-cell memristor chip. By interacting brain, continuously its parameters, leading successful real-time control drone in 4 degrees freedom (4-DOF) and enabling it navigate around obstacles. Our approach featured hardware-efficient one-step strategy, chip-equipped achieve performance equivalent software-based methods. Notably, accomplished at three orders magnitude lower energy consumption two higher normalized speed than central processing unit (CPU). Moreover, employing an interactive update framework, showed co-evolution brain-memristor over extended interaction task involving ten subjects. resulted remarkable enhancement nearly 20%, showcasing substantial potential decoders advancing BCIs. The study results that initially played dominant role co-evolution, learned as process progressed. Eventually, dynamic balance between emerged for decision-making. These findings lay groundwork developing future human-centric hybrid intelligence systems.

Language: Английский

Citations

1
Temporally Local Weighting-Based Phase-Locked Time-Shift Data Augmentation Method for Fast-Calibration SSVEP-BCI DOI Creative Commons
Jiayang Huang, Yidan Lv, Zhiqiang Zhang

et al.

IEEE Transactions on Neural Systems and Rehabilitation Engineering, Journal Year: 2024, Volume and Issue: 32, P. 2376 - 2387

Published: Jan. 1, 2024

Various training-based spatial filtering methods have been proposed to decode steady-state visual evoked potentials (SSVEPs) efficiently. However, these require extensive calibration data obtain valid filters and temporal templates. The time-consuming collection process would reduce the practicality of SSVEP-based brain-computer interfaces (BCIs). Therefore, we propose a temporally local weighting-based phase-locked time-shift (TLW-PLTS) augmentation method augment training for calculating In this method, sliding window strategy using SSVEP response period as step is generate augmented data, time filter which maximises covariance between original template signal sine-cosine reference used suppress noise in data. For performance evaluation, TLW-PLTS was incorporated with state-of-the-art calculate classification accuracies information transfer rates (ITRs) three datasets. Compared other methods, demonstrates superior decoding fewer promising development fast-calibration BCIs.

Language: Английский

Citations

1
Cross Stimulus Transfer Learning Framework Using Common Period Repetition Components for Fast Calibration of SSVEP Based BCIs DOI
Jing Jin, Xinjie He, Ren Xu

et al.

IEEE Internet of Things Journal, Journal Year: 2024, Volume and Issue: 12(5), P. 5719 - 5731

Published: Oct. 31, 2024

Language: Английский

Citations

1
Improving Generalized Zero-Shot Learning SSVEP Classification Performance From Data-Efficient Perspective DOI Creative Commons
Xietian Wang, Aiping Liu, Le Wu

et al.

IEEE Transactions on Neural Systems and Rehabilitation Engineering, Journal Year: 2023, Volume and Issue: 31, P. 4135 - 4145

Published: Jan. 1, 2023

Generalized zero-shot learning (GZSL) has significantly reduced the training requirements for steady-state visual evoked potential (SSVEP) based brain-computer interfaces (BCIs). Traditional methods require complete class data sets training, but GZSL allows only partial sets, dividing them into 'seen' (those with data) and 'unseen' classes without data). However, inefficient utilization of SSVEP limits accuracy information transfer rate (ITR) existing methods. To this end, we proposed a framework more effective at three systematically combined levels: acquisition, feature extraction, decision-making. First, prevalent SSVEP-based BCIs overlook inter-subject variance in latency employ fixed sampling starting time (SST). We introduced dynamic (DSST) strategy acquisition level. This uses classification results on validation set to find optimal (OSST) each subject. In addition, developed Transformer structure capture global input compensating small receptive field networks. The fields can adequately process from longer sequences. For decision-making level, designed classifier selection that automatically select seen unseen classes, respectively. also procedure make above solutions conjunction other. Our method was validated public datasets outperformed state-of-the-art (SOTA) Crucially, representative all classes.

Language: Английский

Citations

3
Almost free of calibration for SSVEP-based brain-computer interfaces DOI
Ruixin Luo, Xiaolin Xiao, Enze Chen

et al.

Journal of Neural Engineering, Journal Year: 2023, Volume and Issue: 20(6), P. 066013 - 066013

Published: Nov. 10, 2023

. Steady-state visual evoked potential (SSVEP)-based brain-computer interface (BCI) is a promising technology that can achieve high information transfer rate (ITR) with supervised algorithms such as ensemble task-related component analysis (eTRCA) and task-discriminant (TDCA). However, training individual models requires tedious time-consuming calibration process, which hinders the real-life use of SSVEP-BCIs. A recent data augmentation method, called source aliasing matrix estimation (SAME), generate new EEG samples from few trials. But SAME does not exploit across stimuli well only reduces number trials per command, so it still has some limitations.

Language: Английский

Citations

3
Phase-Locked Time-Shift Data Augmentation Method for SSVEP Brain-Computer Interfaces DOI Creative Commons
Ximing Mai, Jikun Ai, Yuxuan Wei

et al.

IEEE Transactions on Neural Systems and Rehabilitation Engineering, Journal Year: 2023, Volume and Issue: 31, P. 4096 - 4105

Published: Jan. 1, 2023

Steady-state visual evoked potential (SSVEP) based brain-computer interfaces (BCIs) have achieved an information transfer rate (ITR) of over 300 bits/min, but abundant training data is required. The performance SSVEP algorithms deteriorates greatly under limited data, and the existing time-shift augmentation method fails to improve it because phase-locked requirement between samples violated. To address this issue, study proposes a novel method, namely (PLTS), for SSVEP-BCI. similarity epochs at different time moments was evaluated, unique step calculated each class augment additional in trial. results showed that PLTS significantly improved classification on BETA datasets. Moreover, condition one calibration block, by slightly prolonging duration (from 48 s 51.5 s), ITR increased from 40.88±4.54 bits/min 122.61±7.05 with PLTS. This provides new perspective augmenting training-based SSVEP-BCI, promotes accuracy thus facilitates real-life applications SSVEP-based brain spellers.

Language: Английский

Citations

2
Enhancing One-Shot Ssvep Classification by Combining Cross-Subject Dual-Domain Fusion Network with Task-Related and Task-Discriminant Component Analysis DOI
Yang Deng, Zhiwei Ji, Yijun Wang

et al.

Published: Jan. 1, 2024

Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI

Language: Английский

Citations

0
Shift Aliasing Signal Data Augmentation Method for SSVEP-based BCIs DOI
Jiayang Huang, Haoyu Wen, Pengfei Yang

et al.

2021 27th International Conference on Mechatronics and Machine Vision in Practice (M2VIP), Journal Year: 2024, Volume and Issue: unknown, P. 1 - 6

Published: Oct. 3, 2024

Language: Английский

Citations

0
Enhancing detection of SSVEP-based BCIs via a novel temporally local canonical correlation analysis DOI

Guangshu Xia,

Li Wang,

Shiming Xiong

et al.

Journal of Neuroscience Methods, Journal Year: 2024, Volume and Issue: 414, P. 110325 - 110325

Published: Nov. 20, 2024

Language: Английский

Citations

0
Improving the Performance of Individually Calibrated SSVEP Classification by Rhythmic Entrainment Source Separation DOI Creative Commons
Wei Xu, Yufeng Ke, Dong Ming

et al.

IEEE Transactions on Neural Systems and Rehabilitation Engineering, Journal Year: 2024, Volume and Issue: 32, P. 4284 - 4293

Published: Jan. 1, 2024

The supervised decoding algorithms of Steady-State Visual Evoked Potentials (SSVEP) have achieved remarkable performance with sufficient training data. However, these methods typically failed to achieve acceptable in single-trial scenarios. To address this challenge, we propose a method enhance SSVEP classification using less data by employing Rhythmic Entrainment Source Separation (RESS) construct spatial filters. We evaluate RESS alongside other state-of-the-art two distinct datasets assess their effectiveness. Our results indicate that significantly outperforms advanced when trained single block calibration Specifically, compared task-related component analysis, the RESS-based improves average accuracy 49.81% and 59.06% on 1-second EEG segments. can improve limited holds promise for practical applications SSVEP-based BCIs, offering novel solution reduce requirements individually calibrated systems.

Language: Английский

Citations

0