Affiliations: The College of Foreign Languages, Zhoukou Normal University, Zhoukou, Henan, China
Correspondence:
[*]
Corresponding author. Wang Yipu, The College of Foreign Languages, Zhoukou Normal University, Zhoukou, Henan, 466226, China. E-mail: keaiyobo@hotmail.com.
Abstract: At present, in the context of the Internet of Things (IoT), more and more teaching institutions have integrated Internet technology into the English network teaching system. Therefore, based on the IoT technology, an English speech recognition algorithm is proposed based on the network teaching system, and develops the function of the network teaching system oriented to English speech intonation. In this paper, the function of the platform and the design of the key modules of the system are proposed in detail. During the operation of the recognition model, the collected speech signals are preprocessed, effective speech feature parameters are extracted, and each frame feature parameter is composed into a vector sequence. The different sequences are classified, and the speech intonation system is developed based on the classification results. Then the improved particle filter algorithm is used to further improve the accuracy and speed of English speech system recognition, and the function of English speech network teaching system is optimized based on this. Experiments show that the excerpts from the China Daily website in the system of this article, most of them can be spliced with word primitives, reaching more than 90% of the total number of words in the measured text, reflecting the system’s higher word and segment coverage and high accuracy. The research in this paper has certain theoretical reference value for the construction and optimization of English phonetic system and the further development and use of IoT technology.
Keywords: IoT, English phonetic intonation, network teaching system, function, development
