Heterogeneous-architecture-based real-time language conversion device and method

Abstract

The invention discloses a heterogeneous-architecture-based real-time language conversion device and method; the method comprises: acquiring numbers of user-specified input language and output language, allowing a user to input speech through a device, enabling the device to perform recognizing, displaying on a screen, judging, by the user, whether displayed recognition results are correct or not, manually correcting the incorrect results, enabling the device to perform machine translation based on a database according to the recognition results or input text, and enabling the device to display the translation results on the screen, or output synthetic speech through a speech output device based on a synthetic speech database. The basic language translation databases are added into the device herein; the required database is added to the system every time according to a user's selection for a required translation language, and speech translation can be quickly finished by the aid of concurrent quick processing by an FPGA (field programmable gate array), through high-speed operation of a processor based on the heterogeneous architecture of the device and method.

Description

technical field [0001] The invention belongs to the technical field of language data processing, and in particular relates to a real-time language conversion device and conversion method based on a heterogeneous architecture. Background technique [0002] As the Chinese economy is becoming more and more integrated into the development of economic globalization, there are more and more cross-border and cross-language exchanges. For example, in international conferences, people need to conduct increasingly frequent international exchanges with countries, not just English exchanges. , including Portuguese, German, French, Russian and other languages. Although most people try to switch to English to communicate, the language communication still cannot achieve the desired effect, resulting in difficulties in daily communication. [0003] In view of the above-mentioned deficiencies in use, one is to deploy simultaneous interpreters on site, but this is only suitable for meeting oc...

AI Technical Summary

Problems solved by technology

[0005] To sum up, the problems existing in the existing technology are: when the network is unstable, it is easy to cause users to wait for a delay, affect the user's use effect, and cannot meet the user's personalized needs and experience

Images