The Robot Was Taught To Copy Human Facial Expressions - Alternative View

The Robot Was Taught To Copy Human Facial Expressions - Alternative View
The Robot Was Taught To Copy Human Facial Expressions - Alternative View

Video: The Robot Was Taught To Copy Human Facial Expressions - Alternative View

Video: The Robot Was Taught To Copy Human Facial Expressions - Alternative View
Video: Robots mimics human facial expressions 2024, May
Anonim

A Japanese engineer has created a robot with a human-like head capable of expressing emotions through the movements of the eyes, eyelids and eyebrows. In addition, he knows how to follow a person's facial expressions and repeat it on his face, as well as fix his gaze on a person. The development was presented at the SIGGRAPH 2018 conference.

Since robots are increasingly being offered to be used not in production, but in the everyday life of people, many laboratories and research organizations are studying the problems that arise when a robot comes into contact with a person. One of the main directions in this area is the expression of emotions, which play an important role in human communication. Some engineers are engaged in the software side of the issue and, for example, teach neural networks to determine the facial expressions of the interlocutor and adjust the facial expressions of the robot in response, while others create physical implementations of robots that have facial expressions.

Japanese engineer Takayuki Todo combined the two approaches and created the SEER robot, which can detect human facial expressions and repeat them. The robot is a small model of a human head. The developer has built in the head realistic models of eyes that can rotate in two directions. Above the eyes are the eyelids, which are also driven by motors. Eyebrows are set above the eyelids, each of which is a flexible wire attached on one side to a mechanism that has three degrees of freedom. Thanks to this, the robot can bend its eyebrows to display facial expressions.

The robot works in tandem with a laptop webcam. It recognizes the person closest to the robot and determines the location of parts of its face in real time. The robot can work in two modes: to repeat the facial expressions of a person and to fix the gaze on him, regardless of the position of the head. The developer notes that while the robot lacks a mouth, which also plays an important role in facial expressions. He plans to install a mechanical analogue of the lips in the next version of the robot.