Speech scoring method and device, electronic device, and storage medium

A voice and voice data technology, applied in the computer field, can solve problems such as inaccurate voice scoring, and achieve the effects of improving scoring accuracy, improving voice recognition accuracy, and accurate evaluation scores
CN109256152AInactive Publication Date: 2019-01-22上海一起作业信息科技有限公司

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
上海一起作业信息科技有限公司
Publication Date
2019-01-22
Estimated Expiration
Not applicable · inactive patent

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses a speech scoring method and device, an electronic device, and a storage medium, and relates to the technical field of computers. The speech scoring method includes the steps that sample features are extracted from sample speech data, and an acoustic model is trained through the sample features, and the trained acoustic model is obtained; a language model is constructed according to standard text data corresponding to the sample speech data, the sample speech data are decoded by the language model and the trained acoustic model to obtain acoustic features of the sample speech data; and a scoring model is trained through the acoustic features and prosodic features of the sample speech data, and target speech data is scored according to the trained scoring model to obtain scores of the target speech data. The speech scoring method and device, the electronic device, and the storage medium can accurately score the target speech data.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The present disclosure relates to the field of computer technology, and in particular, to a voice scoring method, a voice scoring device, electronic equipment, and a computer-readable storage medium. Background technique

[0002] With the development of computer technology, students' speech can be automatically recognized and evaluated through the oral evaluation system.

[0003] Most of the speech evaluation systems in the related art are based on adult audio-driven speech acoustic models, and children's speech is quite different from adult speech acoustic models, which leads to inaccurate recognition; in addition, most speech evaluation systems on the market use acoustic models. Score, using the combination of acoustic features, evaluate the final score through linear regression or support vector machine. However, a single linear combination of acoustic scores cannot match the scores of professional grading teachers, and the evaluation efficiency is...

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