文件名称：Language Modeling for Automatic Speech Recognition of Inflective Languages
文件大小：1.41MB
文件格式：PDF
更新时间：2021-05-27 09:49:22
Speech Recognition
In the few last years, we saw the rise of practical speech recognition applications, which work well in English and a few other languages. There is no doubt that this trend will continue and a more natural interaction between humans and technology will become part of our lives. Language is one of the most important components of one’s culture and identity. With English being the most present language in information technology, other languages face the danger of being diminished in this area. The development of speech recognition systems represents a potential risk for further diminishing languages, in which a sufficient speech recognition performance is harder to achieve. This does not consider languages spoken by very few people but also other languages spoken by many people, which are simply hard in speech recognition, like inflective languages, i.e., languages which due to the inflection of main parts of speech have a rich morphology and therefore the need for a much larger vocabulary in speech recognition. This book gives an overview of inflective languages and their place in speech recognition. We describe the most important language features that are important for the development of a speech recognition system. This is then presented through the analysis of errors in the system and the development of language models and their inclusion in speech recognition systems, which specifically address the errors that are relevant for targeted applications.