Speech translation method, system and device, terminal and storage medium

The invention relates to a speech translation method, system and device, a terminal and a storage medium. The speech translation method comprises: training an initial translation system based on M groups of training samples to obtain a trained translation system, M being N + 1, and N being a positiv...

Full description

Saved in:
Bibliographic Details
Main Authors XU ZHEZHE, HUANG WUWEI, LI XIANG
Format Patent
LanguageChinese
English
Published 05.04.2024
Subjects
ACOUSTICS
CALCULATING
COMPUTING
COUNTING
ELECTRIC DIGITAL DATA PROCESSING
MUSICAL INSTRUMENTS
PHYSICS
SPEECH ANALYSIS OR SYNTHESIS
SPEECH OR AUDIO CODING OR DECODING
SPEECH OR VOICE PROCESSING
SPEECH RECOGNITION
Online AccessGet full text

Cover

More Information
Summary:The invention relates to a speech translation method, system and device, a terminal and a storage medium. The speech translation method comprises: training an initial translation system based on M groups of training samples to obtain a trained translation system, M being N + 1, and N being a positive integer greater than 1; a first voice segment in the voice signal is input to a voice encoder, a voice state sequence corresponding to the first voice segment output by the voice encoder is obtained, the voice signal comprises a plurality of voice segments, and the voice signal belongs to a source language; and inputting the voice state sequence corresponding to the first voice segment and the N reference data into a text decoder to respectively obtain target text segments of N target languages corresponding to the first voice segment output by the text decoder. According to the translation system provided by the method, the voice signals of one source language are translated into texts of multiple target languag
Bibliography:Application Number: CN202211199610