Sinc-Windowing and Multiple Correlation Coefficients Improve SSVEP Recognition Based on Canonical Correlation Analysis

Abstract

Canonical Correlation Analysis (CCA) is an increasingly used approach in the field of Steady-State Visually Evoked Potential (SSVEP) recognition. The efficacy of the method has been widely proven, and several variations have been proposed. However, most CCA variations tend to complicate the method, usually requiring additional user training or increasing computational load. Taking simple procedures and low computational costs may be, however, a relevant aspect, especially in view of low-cost and high-portability devices. In addition, it would be desirable that the proposed variations are as general and modular as possible to facilitate the translation of results to different algorithms and setups. In this work, we evaluated the impact of two simple, modular variations of the classical CCA method. The variations involved (i) the number of canonical correlations used for classification and (ii) the inclusion of a prefiltering step by means of sinc-windowing. We tested ten volunteers in a 4-class SSVEP setup. Both variations significantly improved classification accuracy when they were used separately or in conjunction and led to accuracy increments up to 7-8% on average and peak of 25–30%. Additionally, variations had no (variation (i)) or minimal (variation (ii)) impact on the number of algorithm steps required for each classification. Given the modular nature of the proposed variations and their positive impact on classification accuracy, they might be easily included in the design of CCA-based algorithms that are even different from ours.

Brain–computer interfaces based on code-modulated visual evoked potentials (c-VEP): a literature review

“…As explained in section 3.2, the first column of W b is used as spatial filter. However, Mondini et al [104] recently hypothesized that relevant information might be spread over more than one coefficient. For that reason, several studies opted to concatenate the projected responses of n filters w b1 , w , .…”

Asynchronous c-VEP communication tools—efficiency comparison of low-target, multi-target and dictionary-assisted BCI spellers

“…However, due to the noisiness of the EEG, information may be distributed over several coefficients 5 . Recently, Mondini et al 26 showed that considering multiple correlations can improve signal classification.…”

“…The authors tested several different decomposition designs: equally spaced, harmonic and overlapping sub-bands, and observed that the latter yielded the highest accuracy. Recently, Monidini et al 26 investigated the number of correlations coefficients considered for CCA classification and found a significant improvement in classification accuracy if more than one coefficients (as in the conventional approach) were considered.…”

Dynamic time window mechanism for time synchronous VEP-based BCIs—Performance evaluation with a dictionary-supported BCI speller employing SSVEP and c-VEP

“…Typically, only the first canonical correlation and corresponding weight is used for classification and construction of filters [4, 5, 29]. Nonetheless, some recent studies yielded better performance when the additional canonical variable pairs with their associated correlations and weights where also utilized [30].…”