Interaction system, non-transitory computer readable storage medium, and method for controlling interaction system
AssigneeHitachi, Ltd.
InventorsTakashi Numata, Ryuji Mine, Yasuhiro Asa
Life of this patent
Estimating the term…
Abstract
An interaction system that interacts with a user is disclosed. The interaction system includes: an input device that receives a speech signal of the user; a computing device that determines a speech content of the interaction system for a speech content acquired from the speech signal of the user such that a frequency distribution of speech feature values of the speech content of the interaction system approaches an ideal frequency distribution; and an output device that outputs the determined speech content of the interaction system.
Claims11
An interaction system, which performs an interaction with a user, comprising:
- a microphone;
- a speaker;
- a processor coupled to the microphone and the speaker;
- and
- a memory coupled to the processor, the memory storing instructions that configured the processor to:
- enable the microphone to receive a speech signal of the user,
- determine a speech content of the interaction system based on a speech content acquired from the speech signal of the user such that a frequency distribution of speech feature values of the speech content of the interaction system approaches an ideal frequency distribution,
- determine a distribution conversion coefficient from the frequency distribution of the speech feature values of the user based on executing a machine learning algorithm to increase the duration of the interaction, the amount of the interaction, and the change or presence or absence of the action of the user with respect to data used in a prior learning,
- calculate the ideal frequency distribution based on the determined distribution conversion coefficient,
- calculate a distance between the frequency distribution of the speech feature values of the speech content of the interaction system and the ideal frequency distribution, and calculate an ideal speech feature value of the speech content of the interaction system based on the distance, and
- cause the speaker to output the determined speech content of the interaction system.
A non-transitory computer readable storage medium storing instructions executed by a computer that interacts with a user, the instructions cause the computer to perform steps comprising:
- acquiring a speech signal of the user from a microphone;
- determining a speech content of the interaction system for a speech content acquired based on the speech signal of the user such that a frequency distribution of speech feature values of the speech content of the interaction system approaches an ideal frequency distribution;
- determining a distribution conversion coefficient from the frequency distribution of the speech feature values of the user based on executing a machine learning algorithm to increase the duration of the interaction, the amount of the interaction, and the change or presence or absence of the action of the user with respect to data used in a prior learning;
- calculating the ideal frequency distribution based on the determined distribution conversion coefficient;
- calculating a distance between the frequency distribution of the speech feature values of the speech content of the interaction system and the ideal frequency distribution, and calculate an ideal speech feature value of the speech content of the interaction system based on the distance;
- and
- causing a speaker to output the determined speech content of the interaction system to an output device.
A control method of an interaction system, which includes a computer and interacts with a user, the method comprising:
- causing the computer to acquire a speech signal of the user;
- causing the computer to determine a speech content of the interaction system based on for a speech content acquired from the speech signal of the user such that a frequency distribution of speech feature values of the speech content of the interaction system approaches an ideal frequency distribution;
- causing the computer to determine a distribution conversion coefficient from the frequency distribution of the speech feature values of the user based on executing a machine learning algorithm to increase the duration of the interaction, the amount of the interaction, and the change or presence or absence of the action of the user with respect to data used in a prior learning;
- causing the computer to calculate the ideal frequency distribution based on the determined distribution conversion coefficient;
- causing the computer to calculate a distance between the frequency distribution of the speech feature values of the speech content of the interaction system and the ideal frequency distribution, and calculate an ideal speech feature value of the speech content of the interaction system based on the distance;
- and
- causing the computer to cause the speaker to output the determined speech content of the interaction system.
Dependencies are inferred from the claim text.