A training-based speech regeneration approach with cascading mapping models.

•A two-step cascading reconstruction algorithm for converting whispered speech to normal speech based on a training-based approach is proposed.•To lessen the effect of inconsistent spectral features and time alignment between normal and whispered speech, an intermediary layer called “artificial whisper” (or whisperised speech) is introduced.•The results of subjective and objective evaluations are presented to compare the efficiency of the proposed algorithm with other computational methods and Electrolarynx.