Audio exchange method and audio exchange system for linguistic semantics

Abstract

The language and semantic audio exchange method, system and audio coding graphics of the present invention solve the technical problems in the prior art that errors in data response and poor real-time performance are caused by complex semantics during inter-language translation. The method includes using the minimum phoneme sequence to form the phonetic mapping structure of each language, and completing the interlanguage conversion of semantics through each phoneme mapping structure. Using the smallest phoneme that forms the smallest short segment of audio in language composition as the basic data exchange unit for semantic conversion between languages, using the smallest phoneme as the coding basis for data exchange, changes the basic structure of speech recognition, and optimizes the coding of audio content in languages The complexity and accuracy prevent the language audio encoding process from being coupled with complex audio features formed by compound information such as pitch, scale, and range in the language segment, ensuring the speech recognition rate. The mapping structure of speech code and text code formed by the minimum phoneme improves the efficiency of data exchange during language translation.

Description

technical field [0001] The invention relates to the field of information exchange, in particular to a language semantic audio exchange method and an audio exchange system. Background technique [0002] The current language translation is mainly composed of speech recognition, semantic analysis and sentence synthesis. Speech recognition uses a high-sensitivity sensor to extract the audio signal set corresponding to the text in the sentence from the frequency domain or time domain speech signal stream of the original language. Analyze and use hidden Markov model (HMM), self-learning model, artificial neural network (ANN) and other models to identify and quantify the text sequence and semantic meaning in the audio signal set to determine the expression content as much as possible, and sentence synthesis is based on the expression content The recognition and quantification data of the target language form a set of audio signals or a set of word sequences. In this process, affec...

AI Technical Summary

Problems solved by technology

[0004] In view of this, the embodiment of the present invention is committed to providing a language semantic audio exchange method and an audio exchange system to solve the technical problems in the prior art that the semantics are complex during language translation, resulting in poor data response and real-time performance

Images