167 search results - page 7 / 34 » Continuous Audio-Visual Speech Recognition |
ICIP
2002
IEEE
14 years 11 months ago
2002
IEEE
CSL
2007
Springer
13 years 9 months ago
2007
Springer
Humans are able to recognise a word before its acoustic realisation is complete. This in contrast to conventional automatic speech recognition (ASR) systems, which compute the lik...
IPPS
1999
IEEE
14 years 2 months ago
1999
IEEE
This paper presents a parallel phoneme recognition algorithm based on the continuous Hidden Markov Model (HMM). The parallel phoneme recognition algorithm distributes 3-state HMMs...
ESANN
1998
13 years 11 months ago
1998
Abstract. In this paper, we focus on a novel NN/HMM architecture for continuous speech recognition. The architecture incorporates a neural feature extraction to gain more discrimin...
MLMI
2007
Springer
14 years 3 months ago
2007
Springer
In this paper we present a study of automatic speech recognition systems using context-dependent phonemes and graphemes as sub-word units based on the conventional HMM/GMM system a...