Robot behavior control system and method, and robot apparatus

• US 20050240412A1
• Filed: 04/06/2005
• Published: 10/27/2005
• Est. Priority Date: 04/07/2004
• Status: Active Grant

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Fujita, Masahiro
• Granted Patent: US 8,145,492 B2
• Time in Patent Office: Days
• US Class Current: 704/270
• CPC Class Codes: G10L 13/00 Speech synthesis; Text to speech systems
• CPC Class Codes: G10L 15/16 using artificial neural networks
• CPC Class Codes: G10L 2015/025 Phonemes, fenemes or fenones being the recognition units

Abstract

A behavior control system of a robot for learning a phoneme sequence includes a sound inputting device inputting a phoneme sequence, a sound signal learning unit operable to convert the phoneme sequence into a sound synthesis parameter and to learn or evaluate a relationship between a sound synthesis parameter of a phoneme sequence that is generated by the robot and a sound synthesis parameter used for sound imitation, and a sound synthesizer operable to generate a phoneme sequence based on the sound synthesis parameter obtained by the sound signal learning unit.

Claims

1. A behavior control system of a robot for learning a phoneme sequence, comprising:

sound inputting means for inputting a phoneme sequence;

sound signal learning means for converting the phoneme sequence into a sound synthesis parameter and for learning or evaluating a relationship between a sound synthesis parameter of a phoneme sequence that is generated by the robot and a sound synthesis parameter used for sound imitation; and

sound synthesizing means for generating a phoneme sequence based on the sound synthesis parameter obtained by the sound signal learning means.

2. The system according to claim 1, wherein the sound inputting means includes a cochlear filter bank.

3. The system according to claim 1, wherein the sound signal learning means is formed of a feed-forward neural network.

4. The system according to claim 1, further comprising:

internal state generating means for generating an internal state or a body constraint; and

mapping means for mapping the internal state to a sound synthesis parameter.

5. The system according to claim 4, wherein the mapping means inversely estimates an internal state corresponding to an input sound synthesis parameter using an inverse function during mapping.

6. A behavior control system of a robot for determining a behavior to be exhibited based on an internal state and an external environment and performing predictive learning, the robot including driving means for driving the robot in response to a driving instruction, internal state managing means for managing the internal state, and measuring means for measuring an external stimulus or a physical environment, the behavior control system comprising:

predictive learning means for estimating a predicted value of the driving instruction for the robot, the internal state, and the physical environment based on a current observed value concerning the driving instruction for the robot, the internal state, and the physical environment and a current target value concerning the driving instruction for the robot, the internal state, and the physical environment.

7. The system according to claim 6, wherein the predictive learning means is formed of a feed-forward neural network, or a correlation learning unit.

8. The system according to claim 6, wherein the predictive learning means performs predictive learning based on the predicted value and a next observed value of the driving instruction for the robot, the internal state, and the physical environment.

9. The system according to claim 6, wherein the predictive learning means uses a tool used to manipulate the physical environment, a body image that is obtained by manipulating the physical environment using the tool, and the environment.

10. The system according to claim 6, wherein the robot further includes a development scenario having a plurality of learning phases, and moves according to a corresponding learning phase in response to the driving instruction.

11. The system according to claim 10, wherein in each of the learning phases, a factor of a flexible movement realized by the robot or a list of behaviors to be performed is defined at individual stages of learning.

12. The system according to claim 10, wherein the robot controls changes of the learning phases over time or based on deviation or correlation between the predicted value estimated by the predictive learning means and a next observed value of the driving instruction for the robot, the internal state, and the physical environment.

13. A behavior control method of a robot for learning a phoneme sequence, comprising the steps of:

inputting a phoneme sequence;

converting the phoneme sequence into a sound synthesis parameter, and learning or evaluating a relationship between a sound synthesis parameter of a phoneme sequence that is generated by the robot and a sound synthesis parameter used for sound imitation; and

generating a phoneme sequence based on the sound synthesis parameter obtained in the step of learning or evaluating.

14. The method according to claim 13, wherein in the step of inputting a phoneme sequence, cochlear filtering is performed.

15. The method according to claim 13, wherein in the step of learning or evaluating, learning based on a feed-forward neural network is performed.

16. The method according to claim 13, further comprising the steps of:

generating an internal state or a body constraint; and

mapping the internal state to a sound synthesis parameter.

17. The method according to claim 13, further comprising the step of inversely estimating an internal state corresponding to an input sound synthesis parameter using an inverse function during mapping performed in the step of mapping.

18. A behavior control method of a robot for determining a behavior to be exhibited based on an internal state and an external environment and performing predictive learning, the behavior control method comprising the step of:

performing predictive learning by estimating a predicted value of an instruction to drive the robot, an internal state of the robot, and a physical environment based on a current observed value concerning the instruction, the internal state, and the physical environment and a current target value concerning the instruction, the internal state, and the physical environment.

19. The method according to claim 18, wherein in the step of performing predictive learning, learning is performed based on a feed-forward neural network, or correlation learning.

20. The method according to claim 18, wherein the step of performing predictive learning, predictive learning is performed based on the predicted value and a next observed value of the instruction, the internal state, and the physical environment.

21. The method according to claim 18, wherein the step of performing predictive learning, a tool used to manipulate the physical environment, a body image that is obtained by manipulating the physical environment using the tool, and the physical environment are used.

22. The method according to claim 18, wherein the robot includes a development scenario having a plurality of learning phases, and moves according to a corresponding learning phase in response to the instruction.

23. The method according to claim 22, wherein in each of the learning phases, a factor of a flexible movement realized by the robot or a list of behaviors to be performed is defined at individual stages of learning.

24. The method according to claim 22, wherein the robot controls changes of the learning phases over time or based on deviation or correlation between the predicted value estimated in the step of performing predictive learning and a next observed value of the instruction, the internal state, and the physical environment.

25. A robot apparatus for learning a phoneme sequence, comprising:

a plurality of movable sections;

driving means for driving the plurality of movable sections;

driving control means for controlling the driving means;

sound inputting means for inputting a phoneme sequence;

sound signal learning means for converting the phoneme sequence into a sound synthesis parameter and for learning or evaluating a relationship between a sound synthesis parameter of a phoneme sequence that is generated by the robot apparatus and a sound synthesis parameter used for sound imitation; and

sound synthesizing means for generating a phoneme sequence based on the sound synthesis parameter obtained by the sound signal learning means.

26. A robot apparatus for determining a behavior to be exhibited based on an internal state and an external environment and performing predictive learning, the robot apparatus comprising:

a plurality of movable sections;

driving means for driving the plurality of movable sections;

driving control means for issuing a driving instruction to control the driving means;

internal state managing means for managing the internal state;

measuring means for measuring an external stimulus or a physical environment; and

predictive learning means for estimating a predicted value of the driving instruction, the internal state, and the physical environment based on a current observed value concerning the driving instruction, the internal state, and the physical environment and a current target value concerning the driving instruction, the internal state, and the physical environment.

27. A robot apparatus for determining a behavior to be exhibited based on an internal state and an external environment and performing learning, the robot apparatus comprising:

a plurality of movable sections;

driving means for driving the plurality of movable sections;

driving control means for issuing a driving instruction to control the driving means;

measuring means for measuring an external stimulus or a physical environment;

learning means for performing learning based on at least a behavior obtained by the driving means and the physical environment; and

a development scenario having a plurality of learning phases,

wherein the driving means causes a motion according to a corresponding learning phase in response to the driving instruction.

28. A behavior control system of a robot for learning a phoneme sequence, comprising:

a sound inputting device inputting a phoneme sequence;

a sound signal learning unit operable to convert the phoneme sequence into a sound synthesis parameter and to learn or evaluate a relationship between a sound synthesis parameter of a phoneme sequence that is generated by the robot and a sound synthesis parameter used for sound imitation; and

a sound synthesizer operable to generate a phoneme sequence based on the sound synthesis parameter obtained by the sound signal learning unit.

29. A behavior control system of a robot for determining a behavior to be exhibited based on an internal state and an external environment and performing predictive learning, the robot including a driving unit driving the robot in response to a driving instruction, an internal state management unit managing the internal state, and a sensor measuring an external stimulus or a physical environment, the behavior control system comprising:

a predictive learning unit operable to estimate a predicted value of the driving instruction for the robot, the internal state, and the physical environment based on a current observed value concerning the driving instruction for the robot, the internal state, and the physical environment and a current target value concerning the driving instruction for the robot, the internal state, and the physical environment.

30. A robot apparatus for learning a phoneme sequence, comprising:

a plurality of movable sections;

a driving unit driving the plurality of movable sections;

a driving control unit controlling the driving unit;

a sound inputting device inputting a phoneme sequence;

a sound signal learning unit operable to convert the phoneme sequence into a sound synthesis parameter and to learn or evaluate a relationship between a sound synthesis parameter of a phoneme sequence that is generated by the robot apparatus and a sound synthesis parameter used for sound imitation; and

a sound synthesizer operable to generate a phoneme sequence based on the sound synthesis parameter obtained by the sound signal learning unit.

31. A robot apparatus for determining a behavior to be exhibited based on an internal state and an external environment and performing predictive learning, the robot apparatus comprising:

a plurality of movable sections;

a driving unit driving the plurality of movable sections;

a driving control unit operable to issue a driving instruction to control the driving unit;

an internal state management unit managing the internal state;

a sensor measuring an external stimulus or a physical environment; and

a predictive learning unit operable to estimate a predicted value of the driving instruction, the internal state, and the physical environment based on a current observed value concerning the driving instruction, the internal state, and the physical environment and a current target value concerning the driving instruction, the internal state, and the physical environment.

32. A robot apparatus for determining a behavior to be exhibited based on an internal state and an external environment and performing learning, the robot apparatus comprising:

a plurality of movable sections;

a driving unit driving the plurality of movable sections;

a driving control unit operable to issue a driving instruction to control the driving unit;

a sensor measuring an external stimulus or a physical environment;

a learning unit operable to perform learning based on at least a behavior obtained by the driving unit and the physical environment; and

a development scenario having a plurality of learning phases,

wherein the driving unit causes a motion according to a corresponding learning phase in response to the driving instruction.