4.2 | Recognition Accuracy
Figure 6a illustrates the contrast in recognition accuracies between the
SF_CCA method pioneered by Yan et al. [5] and the novel ERLLP_CCA
approach introduced in this research, spanning a data analysis length
from 1 to 2 seconds in 0.2-second increments. Within this framework, the
CCA is applied to the foundational signal X , the SF_CCA is
implemented on the refined filtered signal X’ , and the ERLLP_CCA
focuses on the ERLLP-amplified signal X” . As highlighted in
Figure 6a, irrespective of the stimulus durations, the ERLLP_CCA
showcases a remarkable enhancement in recognition accuracy compared to
the original SSVEP signal, distinctly outperforming the capabilities of
SF_CCA. For stimulus durations at intervals of 1, 1.2, 1.4, 1.6, 1.8,
and 2 seconds, the ERLLP_CCA methodology yields improved accuracy rates
by 3.08%, 4.28%, 2.02%, 2.08%, 2.02%, and 1.08% compared to the
baseline SF_CCA method, respectively.