Titlebook: Speech Recognition Algorithms Using Weighted Finite-State Transducers; Takaaki Hori,Atsushi Nakamura Book 2013 Springer Nature Switzerland

ANA

glomeruli

Takaaki Hori,Atsushi Nakamuraion not only in telecommunication, starting from telephone exchanges through ATM switches and IP routers to optical cross-connect systems and optical packet switches, but also in other areas of knowledge like computation and control. Theory of switching fabrics becomes also a part of applied mathema

食料

不易燃

1932-121X he Weighted Finite-State Transducer (WFST) approach. The decoding process for speech recognition is viewed as a search problem whose goal is to find a sequence of words that best matches an input speech signal. Since this process becomes computationally more expensive as the system vocabulary size i

NATTY

Grievance

Summary and Perspective,oding with the WFST. Finally, we described two dynamic decoders with on-the-fly composition of multiple WFSTs, which are more tractable for practical speech applications than the basic approach with a fully composed WFST.

亲爱

Brief Overview of Speech Recognition,owever, since the technology is wide ranging and progressing day by day, this chapter focuses specifically on a standard approach to large vocabulary continuous speech recognition (LVCSR), which covers continuous-density hidden Markov models, context-dependent phone modeling, .-gram language models,

门窗的侧柱

终端

危机

Summary and Perspective,tals of large-vocabulary continuous-speech recognition (LVCSR). Specifically, we focused on the decoding problem in LVCSR and discussed why the WFST-based approach improved search efficiency in the decoding process. We also presented concrete algorithms for constructing a fully composed WFST and dec