Apparatus, method, and non-transitory computer readable storage medium thereof for generating control instructions based on text

Patent metadata

Field	Value
Publication number	US-10460731-B2
Application number	US-201815936416-A
Country	US
Kind code	B2
Filing date	Mar 26, 2018
Priority date	Nov 30, 2017
Publication date	Oct 29, 2019
Grant date	Oct 29, 2019

Abstract

An apparatus, method, and non-transitory computer readable storage medium for generating control instructions based on text are provided. The apparatus analyzes a text with at least one string to generate a text timestamp document, analyzes the at least one string to generate a plurality of pieces of emotion information, a plurality of pieces of action information, and a plurality of pieces of environment information, and decides a plurality of emotion tags and their corresponding timestamps, a plurality of action tags and their corresponding timestamps, and a plurality of environment tags and their corresponding timestamps according to a feature database. The apparatus further generates control instruction(s) and the corresponding timestamp(s) according to the emotion tags and their corresponding timestamps, the action tags and their corresponding timestamps, and the environment tags and their corresponding timestamps.

First claim

What is claimed is:

1. An apparatus for generating control instructions for playing a text, comprising:
an input interface, being configured to receive the text comprising at least one string;
a storage, being configured to store a feature database and a control instruction set, wherein the feature database comprises a plurality of preset emotion tags, a plurality of preset action tags, and a plurality of preset environment tags, the control instruction set comprises a plurality of control instructions, and each of the control instructions corresponds to at least one of the preset emotion tags, the preset action tags, and the preset environment tags; and
a processor electrically connected to the input interface and the storage, being configured to analyze the at least one string comprised in the text to generate a text timestamp datum, wherein the text timestamp datum comprises at least one time for playing the at least one string respectively, and the processor further analyzes the at least one string to obtain a plurality of pieces of emotion information, a plurality of pieces of action information, and a plurality of pieces of environment information, decide a plurality of first emotion tags and corresponding timestamps thereof, a plurality of action tags and corresponding timestamps thereof, and a plurality of environment tags and corresponding timestamps thereof for the text according to the feature database and the text timestamp datum, and generate at least one first control instruction and corresponding at least one timestamp thereof for playing the text according to the first emotion tags and the corresponding timestamps thereof, the action tags and the corresponding timestamps thereof, and the environment tags and the corresponding timestamps thereof,
wherein each of the first emotion tags is one of the preset emotion tags, each of the action tags is one of the preset action tags, and each of the environment tags is one of the preset environment tags.

2. The apparatus for generating control instructions for playing a text of claim 1 , wherein the text records at least one role, each of the first emotion tags corresponds to one of the at least one role, and each of the action tags corresponds to one of the at least one role.

3. The apparatus for generating control instructions for playing a text of claim 2 , wherein the processor further calculates an emotion correlation degree between the first emotion tags and the corresponding timestamps thereof, the action tags and the corresponding timestamps thereof, and the environment tags and the corresponding timestamps thereof corresponds to one of the at least one role, and the processor further corrects the first emotion tags into a plurality of second emotion tags according to the emotion correlation degree, each of the second emotion tags is one of the preset emotion tags, and the processor generates the at least one first control instruction and the corresponding at least one timestamp thereof according to the second emotion tags and the corresponding timestamps thereof, the action tags and the corresponding timestamps thereof, and the environment tags and the corresponding timestamps thereof.

4. The apparatus for generating control instructions for playing a text of claim 3 , wherein the processor further calculates an instruction sequence text datum according to the text timestamp datum, each of the second emotion tags and the corresponding timestamp thereof, each of the action tags and the corresponding timestamp thereof, each of the environment tags and the corresponding timestamp thereof, and the control instruction set, wherein the instruction sequence text datum comprises a plurality of second control instructions corresponding to the second emotion tags respectively and a plurality of corresponding timestamps thereof, a plurality of third control instructions corresponding to the action tags respectively and a plurality of corresponding timestamps thereof, and a plurality of fourth control instructions corresponding to the environment tags respectively and a plurality of corresponding timestamps thereof, and the storage further stores the instruction sequence text datum.

5. The apparatus for generating control instructions for playing a text of claim 1 , further comprising:
a voice input interface electrically connected to the processor, being configured to receive a voice message,
wherein the processor further recognizes the voice message to obtain the text.

6. The apparatus for generating control instructions for playing a text of claim 4 , wherein each of the second control instructions, the third control instructions, and the fourth control instructions comprises a to-be-controlled device datum and an execution instruction.

7. The apparatus for generating control instructions for playing a text of claim 1 , wherein the processor further modifies one of the at least one first control instruction and the corresponding timestamp thereof according to a response signal.

8. A method for generating control instructions for playing a text, being adapted for an electronic computing apparatus, and the method comprising:
(a) analyzing at least one string comprised in the text to generate a text timestamp datum, wherein the text timestamp datum comprises at least one time for playing the at least one string respectively;
(b) analyzing the at least one string comprised in the text to obtain a plurality of pieces of emotion information, a plurality of pieces of action information, and a plurality of pieces of environment information;
(c) deciding a plurality of first emotion tags and corresponding timestamps thereof, a plurality of action tags and corresponding timestamps thereof, and a plurality of environment tags and corresponding timestamps thereof for the text according to a feature database and the text timestamp datum; and
(d) generating at least one first control instruction and corresponding at least one timestamp thereof for playing the text according to the first emotion tags and the corresponding timestamps thereof, the action tags and the corresponding timestamps thereof, and the environment tags and the corresponding timestamps thereof.

9. The method for generating control instructions for playing a text of claim 8 , wherein the text records at least one role, each of the first emotion tags corresponds to one of the at least one role, and each of the action tags corresponds to one of the at least one role.

10. The method for generating control instructions for playing a text of claim 9 , wherein the step (d) comprises:
calculating an emotion correlation degree between the first emotion tags and the corresponding timestamps thereof, the action tags and the corresponding timestamps thereof, the environment tags and the corresponding timestamps thereof, and the at least one role;
correcting the first emotion tags into a plurality of second emotion tags according to the emotion correlation degree; and
generating the at least one first control instruction and the corresponding at least one timestamp thereof according to the second emotion tags and the corresponding timestamps thereof, the action tags and the corresponding timestamps thereof, and the environment tags and the corresponding timestamps thereof.

11. The method for generating control instructions for playing a text of claim 10 , further comprising:
calculating an instruction sequence text datum according to the text timestamp datum, each of the second emotion tags and the corresponding timestamp thereof, each of the action tags and the corresponding timestamp thereof, each of the environment tags and the corresponding timestamp thereof, and the control instruction set, wher

Assignees

• Inst Information Ind

Inventors

• Chen Pin-Liang
• Yang Ping-Che
• Ku Tsun

Classifications

• G06F40/30 Primary

  Semantic analysis · CPC title

• G10L15/00

  Speech recognition (G10L17/00 takes precedence) · CPC title

• G06F3/167

  Audio in a user interface, e.g. using voice commands for navigating, audio feedback · CPC title

• G10L15/26 Primary

  Speech to text systems (G10L15/08 takes precedence) · CPC title

• G10L2015/223

  Execution procedure of a spoken command · CPC title