Speech synthesis system based on big data and evaluation of Japanese language feeling

Abstract

With the progress and development of modern technology, people have put forward higher standards for speech synthesis technology. Because speech synthesis technology has already been applied to various fields and industries in people’s lives at this stage, speech synthesis technology faces such high standards. Some are overwhelmed. The main problem now is that the synthesized speech is not natural enough compared to human speech, the pronunciation is not standard enough, and the emotional color is lacking. The main basis of this paper is the continuous self-learning and optimized speech synthesis technology in the Markov model, analyzes its main algorithms, and mainly optimizes its input text analysis and output speech synthesis algorithms. Language sense is a kind of resonance with a language when it is used for a long time, so it is basically based on the proficiency of a language. Therefore, the quality of a language can directly reflect the proficiency of a language, and it is also an important guarantee for smooth communication. Among them, Japanese is an important medium for carrying Japanese culture. The study of Japanese vocabulary will play a very important role in the study of Japanese language sense, and it has certain significance. In daily Japanese learning, consciously paying attention to language sense can accelerate the formation of language sense. Therefore, when you have a better sense of language, you can quickly and directly resonate with the language. So in the process of learning Japanese, if you can have a better sense of Japanese, then Japanese learning will advance by leaps and bounds. Therefore, this paper studies the speech synthesis system and Japanese language sense and its evaluation.