Many techniques for complex speech processing such as denoising and deconvolution, time/frequency warping, multiple speaker separation, and multiple microphone analysis operate on...
A recent area of significant progress in speaker recognition is the use of high level features—idiolect, phonetic relations, prosody, discourse structure, etc. A speaker not on...
William M. Campbell, Joseph P. Campbell, Douglas A...
Manyimportant physical phenomena,such as temperature distribution, air flow, and acoustic waves,are describedas continuous,distributed parameterfields. Analyzingandcontrolling the...
This paperconsidersthe problem of representingcomplex systems that evolve stochastically over time. Dynamic Bayesian networks provide a compact representation for stochastic proce...
Word sense disambiguation for unrestricted text is one of the most difficult tasks in the fields of computational linguistics. The crux of the problem is to discover a model that ...