Research on the Translation System of Spoken English Situational Dialogue Based on Computer Artificial Intelligence

In this paper, a sequence-to-sequence model is proposed to deal with continuous emphasis levels, and machine translation and emphasis translation are combined into a spoken English situational translation system. In the aspect of hardware design, the peripheral circuit is built with the chip as the core to form the speech recognition module. In the process of software design, relaxation endpoint method is used to improve the operation speed to realize floating point operation in the kernel. A database suitable for spoken translation system is constructed by using complex feature vectors, and remote data synchronization is realized by using SQL Server Mobile 2015 database merge replication technology. The system uses a translation engine to translate English and Chinese. There are nearly 200,000 bilingual sentences debugging in the system, which improves the language translation quality of the system.