Method and apparatus for broadcasting a response based on artificial intelligence, and storage medium

Patent metadata

Field	Value
Publication number	US-10923102-B2
Application number	US-201815991411-A
Country	US
Kind code	B2
Filing date	May 29, 2018
Priority date	Jun 22, 2017
Publication date	Feb 16, 2021
Grant date	Feb 16, 2021

Abstract

The present disclosure provides a method and apparatus for broadcasting a response based on artificial intelligence, and a storage medium, wherein the method comprises: obtaining a user-input speech query; generating a response corresponding to the query; obtaining a recorded speech of a mood meaning corresponding to a modal particle in the response and matched with the response; combining the obtained recorded speech with a TTS-generated speech to perform TTS broadcast of the response. The solution of the present disclosure may be applied to enhance an effect of broadcasting the response.

First claim

What is claimed is:

1. A method for broadcasting a response based on artificial intelligence, wherein the method comprises:
obtaining a user-input speech query;
generating a response in a text form corresponding to the query;
in response to determining that the response includes a modal particle, determining a mood meaning of the modal particle expressed in the response, and and obtaining a pre-recorded speech of the modal particle having the mood meaning; and
combining the obtained pre-recorded speech with a Text-To-Speech-generated speech to perform Text-To-Speech broadcast of the response.

2. The method according to claim 1 , wherein
the generating a response corresponding to the query comprises:
determining a demand corresponding to the query;
selecting one response-generating algorithm from M response-generating algorithms corresponding to the demand, M being a positive integer larger than 1; and
using the selected response-generating algorithm to generate the response.

3. The method according to claim 2 , wherein
the determining a demand corresponding to the query comprises:
performing speech recognition for the query to obtain a speech recognition result; and
determining a demand corresponding to the query by performing semantic parsing for the speech recognition result;

the selecting one response-generating algorithm from M response-generating algorithms corresponding to the demand comprises:
randomly selecting one response-generating algorithm from the M response-generating algorithms corresponding to the demand.

4. The method according to claim 1 , wherein
at least one pre-recorded speech is pre-generated for the modal particle, and each pre-recorded speech corresponds to a different mood meaning.

5. The method according to claim 1 , wherein
the combining the obtained pre-recorded speech with a Text-To-Speech-generated speech to perform Text-To-Speech broadcast of the response comprises:
when it is needed to broadcast the modal particle corresponding to the obtained pre-recorded speech, broadcasting the obtained pre-recorded speech, otherwise broadcasting the Text-To- Speech-generated speech.

6. A computer device, comprising a memory, a processor and a computer program which is stored on the memory and runs on the processor, wherein the processor, upon executing the program, implements the following operation:
obtaining a user-input speech query;
generating a response in a text form corresponding to the query;
in response to determining that the response includes a modal particle, determining a mood meaning of the modal particle expressed in the response, and obtaining a pre-recorded speech of the mood meaning having the mood meaning; and
combining the obtained pre-recorded speech with a Text-To-Speech-generated speech to perform Text-To-Speech broadcast of the response.

7. The computer device according to claim 6 , wherein
the generating a response corresponding to the query comprises:
determining a demand corresponding to the query;
selecting one response-generating algorithm from M response-generating algorithms corresponding to the demand, M being a positive integer larger than 1; and
using the selected response-generating algorithm to generate the response.

8. The computer device according to claim 7 , wherein
the determining a demand corresponding to the query comprises:
performing speech recognition for the query to obtain a speech recognition result; and
determining a demand corresponding to the query by performing semantic parsing for the speech recognition result;

the selecting one response-generating algorithm from M response-generating algorithms corresponding to the demand comprises:
randomly selecting one response-generating algorithm from the M response-generating algorithms corresponding to the demand.

9. The computer device according to claim 6 , wherein
pre-recorded speech is pre-generated for the modal particle, and each pre-recorded speech corresponds to a different mood meaning.

10. The computer device according to claim 6 , wherein
the combining the obtained pre-recorded speech with a Text-To-Speech-generated speech to perform Text-To-Speech broadcast of the response comprises:
when it is needed to broadcast the modal particle corresponding to the obtained pre- recorded speech, broadcasting the obtained pre-recorded speech, otherwise broadcasting the Text- To-Speech-generated speech.

11. A non-transitory computer-readable storage medium on which a computer program is stored, wherein the program, when executed by a processor, implements the following operation:
obtaining a user-input speech query;
generating a response in a text form corresponding to the query;
in response to determining that the response includes a modal particle, determining a mood meaning of the modal particle expressed in the response, and obtaining a pre-recorded speech of the modal particle having the mood meaning; and
combining the obtained pre-recorded speech with a Text-To-Speech-generated speech to perform Text-To-Speech broadcast of the response.

12. The non-transitory computer-readable storage medium according to claim 11 , wherein
the generating a response corresponding to the query comprises:
determining a demand corresponding to the query;
selecting one response-generating algorithm from M response-generating algorithms corresponding to the demand, M being a positive integer larger than 1; and
using the selected response-generating algorithm to generate the response.

13. The non-transitory computer-readable storage medium according to claim 12 , wherein
the determining a demand corresponding to the query comprises:
performing speech recognition for the query to obtain a speech recognition result; and
determining a demand corresponding to the query by performing semantic parsing for the speech recognition result;

the selecting one response-generating algorithm from M response-generating algorithms corresponding to the demand comprises:
randomly selecting one response-generating algorithm from the M response-generating algorithms corresponding to the demand.

14. The non-transitory computer-readable storage medium according to claim 11 , wherein
at least one pre-recorded speech is pre-generated for the modal particle, and each pre-recorded speech corresponds to a different mood meaning.

15. The non-transitory computer-readable storage medium according to claim 11 , wherein
the combining the obtained pre-recorded speech with a Text-To-Speech-generated speech to perform Text-To-Speech broadcast of the response comprises:
when it is needed to broadcast the modal particle corresponding to the obtained pre-recorded speech, broadcasting the obtained pre-recorded speech, otherwise broadcasting the Text-To-Speech-generated speech.

Assignees

• Baidu online network technology beijing co ltd

Inventors

• Wang Yu
• Xie Bo

Classifications

• G10L15/22

  Procedures used during a speech recognition process, e.g. man-machine dialogue · CPC title

• G06F16/3344

  using natural language analysis · CPC title

• G10L2015/223

  Execution procedure of a spoken command · CPC title

• G10L25/63

  for estimating an emotional state · CPC title

• G06F40/30

  Semantic analysis · CPC title