A Language Recognition Method Based on Multi-task
A language recognition, multilingual technology, applied in speech recognition, speech analysis, instruments, etc., can solve the problem that it is difficult for models to extract deep speech features, saving time and space costs, improving efficiency, and reducing labor costs.
Active Publication Date: 2022-04-12
北京中科智加科技有限公司
View PDF14 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
In this type of method, the quality of the model often depends on the quality and quantity of the labeled data. In addition, it is difficult for these models to extract deep features of speech, such as the speaker's gender, age, and pronunciation differences between different languages, etc.
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0055] Preferred embodiments of the present invention will be specifically described below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of the application and are used together with the embodiments of the present invention to explain the principle of the present invention.
[0056] A multi-task-based language recognition method in this embodiment, such as figure 1 shown, including the following steps:
[0057] Step S101, building and training a multilingual recognition model based on multi-task learning; the tasks include language recognition tasks and whether it is an effective speech recognition task;
[0058] Step S102, acquiring the voice data to be recognized, and performing preprocessing to obtain the first voice queue to be recognized;
[0059] Step S103, processing the voice data in the first voice queue in batches to obtain multiple batches of second voice queues with similar data lengths in each batch;
[0060]...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention relates to a multi-task-based language recognition method, comprising the following steps: constructing and training a multi-lingual recognition model based on multi-task learning; the task includes a language recognition task and whether it is an effective speech recognition task; and obtaining speech data to be recognized , carry out preprocessing to obtain the first voice queue to be recognized; carry out batch processing of voice data in the first voice queue to obtain multiple batches and second voice queues with similar data lengths in each batch; The second voice queue is sequentially imported into the multilingual recognition model, and after the language recognition and valid voice recognition are performed simultaneously in the language recognition model, the recognition results are output batch by batch. The invention realizes simultaneous recognition of two tasks of language recognition and effective voice recognition, and improves the accuracy and recognition efficiency of language recognition.
Description
technical field [0001] The invention relates to the technical field of speech recognition, in particular to a multi-task-based language recognition method. Background technique [0002] In recent years, speech recognition technology has made significant progress. Language recognition, as one of the important technologies of speech recognition, plays an important role in many fields such as military, medical, and education; The upstream tasks in the system play an important role in improving the accuracy of multilingual language recognition and the user experience of the system; [0003] Traditional language recognition methods usually include three processes: speech signal feature extraction, language model construction, and decision rules. Commonly used speech features in this process include MFCC, Fbank, spectrogram, i-vector, etc. The language model is usually a or multiple classifiers. Commonly used classifiers include SVM, decision tree, random forest, etc. The decisio...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G10L15/00G10L25/87G10L25/90G10L15/06
CPCG10L15/005G10L25/87G10L25/90G10L15/063
Inventor 陈玮冯少辉张建业
Owner 北京中科智加科技有限公司