An online noun phrase translation method based on speech recognition technology

Due to the low translation accuracy of traditional methods, a noun phrase online translation method based on speech recognition technology is proposed. Firstly, an online speech signal recogniser is used to collect the speech signals of noun phrases, and Fourier transform is used for the denoising process. Secondly, based on the denoised speech signal, a benchmark translation context is set to extract the features of noun phrase speech signals in the optimal translation context. Finally, a transformation layer is introduced into the seq2seq model, with the source noun phrase as input and the target noun phrase as output, to construct a neural machine translation model for noun phrases and complete online translation of noun phrases. The experimental results show that the method proposed in this paper can accurately recognise the speech signals of noun phrases and improve the accuracy of online translation. The accuracy of online translation remains above 93%.