新工科视域下语音识别声学模型的设计实验

针对新工科视域下工程专业对学生具有扎实理论基础,较强的实验实践操作能力,强烈工程创新意识的需求,探索抽象理论模型与具象实验相结合的教学模式。设计隐马尔科夫模型的实验中,内容涵盖了数学与统计学、声学与语言学、计算机与人工智能等多学科交叉的知识,通过逐步优化多参数融合的隐马尔科夫模型实验,锻炼了学生理论与实践相结合,分析和解决复杂问题的能力,也解决了学生参与工程实践与科研实践的难题,将知识系统化、工程化、科研化,实现了新工科的培养理念。

In view of the needs of engineering majors in the field of new engineering, students have a solid theoretical foundation, strong experimental practice and engineering innovation consciousness, this paper explores a teaching mode that combines abstract theoretical models with concrete experiments. In the experiment of designing the hidden Markov model, the content covers the interdisciplinary knowledge of mathematics and statistics, acoustics and linguistics, computer and artificial intelligence. The Hidden Markov Model experiment with the gradual optimization of multi-parameter fusion exercises students’ ability to combine theory and practice, to analyze and solve complex problems, also solves the problem of students participating in engineering practice and scientific research practice, systemizing, engineering, thus systematizing, engineering and scientific research knowledge and realizing the cultivation concept of new engineering.