ICPR
2000
IEEE
15 years 1 months ago
2000
IEEE
Speech can be represented as a time/frequency distribution of energy using a multi-band filter bank. A Markov random field model, which takes into account the possible time asynch...