background and aim: speech understanding almost never occurs in silence. verbal communication often occurs in environments where multiple speakers are talking. in such environments, babbling noise masks speech comprehension. consonants, in comparison to vowels, are more sensitive to noise masking. consonants provide most acoustic information needed for comprehending the meaning of the word. sin...

To extract speech from acoustic interference is a challenging problem. Previous systems based on auditory scene analysis principles deal with voiced speech, but cannot separate unvoiced speech. We propose a novel method to separate stop consonants, which contain significant unvoiced signals, based on their acoustic properties. The method employs onset as the major grouping cue; it first detects...

Time-Delay Neural Networks (TDNN) have been shown by Waibel et al. [1] to be a good method for the classification of dynamic speech sounds such as voiced stop consonants. In this paper we discuss key issues in the design and training of a TDNN, based on a Multi-Layer Perceptron (MLP), when used for classification of the sets of voiced stop consonants (/b/, /d/, and /g/) and unvoiced stop conson...

The bursts and voiced formant transitions are well known as separate cues to the place of articulation of initial stop consonant. The Vietnamese presents three final voiceless stop consonants /p, t, k/ without bursts. It is an opportunity to study these final stop consonants and to compare their characteristics with those of the corresponding initial stop consonants. This paper analyses these f...