Graphics Reference
In-Depth Information
2
The Analysis of Speech Recognition Software
If we want to construct speech plotting system, first we should confirm which speech
recognition software is used, we should focus on the recognition rate and the difficul-
ty of the speech recognition software. At present, the mainstream speech recognition
software comprises Nuance, Viavoice, SAPI of Microsoft and the open source soft-
ware HTK, these types of software are continuous speech recognition system for
independent speaker, they have been widely used in the world. Some domestic speech
recognition system is realized based on these types of software [2] [3] [4].
The SAPI of Microsoft is used commonly, at present the operating system Micro-
soft released comprises the speech recognition function, it can realize the text-speech
conversion, but the recognition rate of the speech recognition software needs to be
improved; Nuance is currently widely used in law and medicine, but it is difficult to
develop, the advantage of the software is the high recognition rate [5]; HTK was de-
veloped by the University of Cambridge, the advantages of this software is open
source, but it is also difficult to develop; Viavoice is developed by the IBM, the soft-
ware has a high recognition rate, at present the software has been used in PDA and the
intelligent vehicles, the software provides the interface of development, it can be
embedded into a system, and it is not difficult to develop.
In order to construct the intelligent speech plotting system, we should select the
mature speech recognition software. From the consideration, we choose the IBM
Viavoice as the underlying platform for speech recognition.
3
The Structure of the Intelligent Plotting System
The basic structure of intelligent speech plotting system is shown below.
Fig. 1. The structure of intelligent plotting system
 
Search WWH ::




Custom Search