467 search results - page 91 / 94 » Phoneme segmentation of speech |
INTERSPEECH
2010
13 years 2 months ago
2010
Reconstructing the full contour of the tongue from the position of 3 to 4 landmarks on it is useful in articulatory speech work. This can be done with submillimetric accuracy usin...
TASLP
2010
13 years 2 months ago
2010
We present a scalable medium bit-rate wide-band audio coding technique based on frequency domain linear prediction (FDLP). FDLP is an efficient method for representing the long-ter...
ICASSP
2011
IEEE
12 years 11 months ago
2011
IEEE
In supervector UBM/GMM paradigm, each acoustic file is represented by the mean parameters of a GMM model. This supervector space is used as a data representation space, which has...
ICASSP
2011
IEEE
12 years 11 months ago
2011
IEEE
The use of visual information derived from accurate lip extraction, can provide features invariant to noise perturbation for speech recognition systems and can be also used in a w...
ICASSP
2011
IEEE
12 years 11 months ago
2011
IEEE
To harvest the potential of multi-channel noise reduction methods, it is crucial to have an accurate estimate of the noise correlation matrix. Existing algorithms either assume sp...