A Training Method for Domain Adaptive Acoustic Models
An acoustic model and training method technology, applied in speech analysis, speech recognition, instruments, etc., can solve problems such as performance degradation, large channel interference, insufficient training data, etc., and achieve good recognition performance
Active Publication Date: 2022-02-18
出门问问(苏州)信息科技有限公司
View PDF5 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, in the telephone customer service scenario, due to the low sampling rate, large channel interference and insufficient training data, the overall recognition rate can only reach 50%-70% of 16K speech recognition
In addition, in the commercial application of telephone customer service, users often pay attention to the recognition rate of a specific field. When the training data of the recognition system does not match the application field, the performance will drop significantly, resulting in the general telephone customer service voice recognition system often being unable to perform in these areas. domain use
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
[0071] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided for more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.
[0072] figure 1 is a block diagram of an example computing device 100 arranged to implement a method of training a domain-adaptive acoustic model according to the present disclosure. In a basic configuration 102 , computing device 100 typically includes system memory 106 and one or more processors 104 . A memory bus 108 may be used for communication between the processor 104 and the system memory 106 .
[0073] Depending on the d...
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
Embodiments of the present disclosure provide a domain-adaptive acoustic model training method, device, readable storage medium, and computing device for constructing an acoustic model with excellent recognition effect in a specified domain. The method includes: acquiring voice data of a plurality of designated fields in a similar field and text data corresponding to the voice data of a plurality of designated fields in a similar field; The text data corresponding to the speech data in the adjacent field of the specified field is used to train the acoustic model to obtain a general acoustic model; the speech data in the specified field and the text data corresponding to the speech data in the specified field are obtained; according to the speech data in the specified field The text data corresponding to the speech data of the specified domain is used to train the general acoustic model to obtain a domain adaptive acoustic model.
Description
technical field [0001] The present disclosure relates to the technical field of speech processing, and in particular to a training method, device, readable storage medium and computing equipment for a domain-adaptive acoustic model. Background technique [0002] Automatic speech recognition technology (Automatic Speech Recognition, ASR) has made great progress in the past few years, and in some scenarios, such as voice intelligent assistants, the performance of the most advanced recognition systems has approached human performance. However, in the telephone customer service scenario, due to the low sampling rate, large channel interference and insufficient training data, the overall recognition rate can only reach 50%-70% of the 16K speech recognition level. In addition, in the commercial application of telephone customer service, users often pay attention to the recognition rate of a specific field. When the training data of the recognition system does not match the applica...
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): G10L15/06G10L15/14G10L15/16
CPCG10L15/063G10L15/144G10L15/16
Inventor 钟利民
Owner 出门问问(苏州)信息科技有限公司