期刊名称：Romanian Conference on Human-Computer Interaction
印刷版ISSN：2344-1690
出版年度：2014
期号：3888
页码：51-54
出版社：Matrix ROM
摘要：This paper presents two different applications implementing two different methods for automatic language identification with the purpose of speech synthesis. Implemented algorithms are described, as well as specific working conditions of such applications and applications’ user interfaces are described. While the language identification based on character Unicode encoding detection need few system resources and is very fast, the statistical analysis based algorithms require more resources and are slower. Finding an optimal compromise is both a challenge for developers as for the user, who should decide at runtime the optimal application configuration for real working conditions. At the end of the paper several considerations are presented, concerning the indispensable inertia in language switching if the application is used for differentiated speech synthesis.
