Abstract: An apparatus and method capable of accurately estimating a participant's muscle state are provided. In order to solve this problem, a muscle state estimation device 1 for estimating the state of a participant's muscles is provided. The muscle state estimation device 1 comprises a circumference sensor 2 that measures the circumference of the muscle, a myopotential sensor 3 that measures the electric potential generated by the activity of the muscle and a muscle state estimation unit 4 that calculates the muscle state such that an evaluation function, of which variables are the deviation of the circumference measured by the circumference sensor 2 from the circumference model that defines the relationship between the muscle state and the circumference and the deviation of the electric potential measured by the myopotential sensor 3 from the muscle tension model that defines the relationship between the muscle state and the electric potential, is minimized.

Abstract: A robot behavior generation method includes: generating a human body model; generating a robot model; creating associations between the human body model and the robot model; acquiring first motion data indicating movement of a human body model; selectively choosing movement feature indices having first unknown motion data that indicates movement of the human body model, and setting an evaluation standard for a reproduction error with respect to the normative motion data; selecting a constraint necessary for a robot to move, included in second unknown motion data that indicates the movement of the robot model; calculating first unknown motion data and second unknown motion data the reproduction errors of which based on the evaluation standard are minimum under the association and the constraint; and using the second unknown motion data to control the robot.

Abstract: A robot behavior generation method includes: generating a human body model; generating a robot model; creating associations between the human body model and the robot model; acquiring first motion data indicating movement of a human body model; selectively choosing movement feature indices having first unknown motion data that indicates movement of the human body model, and setting an evaluation standard for a reproduction error with respect to the normative motion data; selecting a constraint necessary for a robot to move, included in second unknown motion data that indicates the movement of the robot model; calculating first unknown motion data and second unknown motion data the reproduction errors of which based on the evaluation standard are minimum under the association and the constraint; and using the second unknown motion data to control the robot.