A robust RNN-based pre-classification for noisy Mandarin speech recognition


This paper addressed the problem of speech signal preclassification for robust noisy speech recognition. A novel RNN-based pre-classification scheme for noisy Mandarin speech recognition is proposed. The RNN, which is trained to be insensitive to noise-level variation, is employed to classify each input frame into the three broad classes of initial, final… (More)


4 Figures and Tables

Slides referencing similar topics