Improving speech communication in the age of face masks: A study on EMD denoising method by subjective speech comparison

•Implementation of the EMD Denoising Speech Enhancement method to enhance single-channel speech signals under various face mask conditions.•Effective utilization of the Fast Fourier Transform (FFT) on the intrinsic mode functions (IMFs) to differentiate between noise and signal during the speech enhancement process.•Validation of the effectiveness of the EMD denoising speech enhancement method through comparison mean opinion scores (CMOS) with a panel of 30 listeners, providing subjective evaluations of the enhanced speech quality. The widespread use of face masks during the COVID-19 pandemic has posed challenges to effective communication. This study investigates the application of the Empirical Mode Decomposition (EMD) denoising method to enhance the quality of acoustic signals affected by face masks. The EMD technique utilizes a Fast Fourier transform (FFT) of the intrinsic mode function to distinguish noise from the signal and improve the acoustic signal. The EMD-enhanced speech signal is compared to the acquired speech signal using Comparison Mean Opinion Scores (CMOS), and the spectral subtraction speech enhancement signal is also evaluated using CMOS. The results from various experimental conditions consistently demonstrate the superior performance of EMD denoising in producing enhanced signals. Specifically, the EMD method outperforms the spectral subtraction method in improving speech signal quality affected by face masks. These findings underscore the effectiveness of the EMD denoising method in improving the quality of speech signals impacted by face masks, highlighting its potential for enhancing communication in challenging acoustic environments.