Robust Speech Recognition in Sports Competition Review based on Natural Language Processing

Abstract Mass communication media is developing at a fast speed, and they have obtained richer methods through different ways of combining, which has attracted people's attention even more. There are many steps in sports competition review, and this article is to study these necessary steps. Sports commentary seems easy, but it is not simple. It includes many links, the content of the explanation should be substantial, and there are many requirements for the language, and at the same time, the emotions of the explanation must be grasped. Every link has a requirement. After summarizing experience, this chapter will rationally use the theory of natural language processing and another theory of robust speech recognition, so as to master each basic link, explain it, and understand the basic laws necessary for review . This article comprehensively describes the robust speech recognition technology, classifies and summarizes several aspects, and conducts more detailed research and discussion on improving its performance. Robust speech recognition systems will always have some problems about mismatch, which requires continuous innovation of some new acoustic models, so that they can be more adapted to the environment. In order to make better use of robust speech recognition performance, in this article, it has innovated the characteristics of its speech, and optimized the parameters and the front-end and back-end cooperation. In this way, it can be better with The combined use of natural language allows commentators to innovate better and increase the popularity of sports competition reviews.