Scientists presented Emo robot, capable of predicting and simultaneously repeating the facial expressions of a person. Development marks significant progress in the field of non -verbal communication of robots.
for more than 5 years, a team of specialists of the Colombian Engineering University worked on the creation of EMO – a robot with a humanoid person equipped 26 drives that allow the robot to reproduce a wide range of facial expressions. Covered in soft silicone skin, EMO’s face is easily configured and served thanks to the magnetic mount system.
The key achievement was EMO training to anticipate a person’s smile 840 milliseconds before it, allowing a robot and a person to smile at the same time. Such interaction erases the line between artificial and natural communication, making the process more natural and genuine.
To achieve this level of interaction, the team developed two models of artificial intelligence: one to predict human facial expressions by analyzing the smallest changes in the person, and the other to generate motor commands corresponding to these expressions. The EMO learning process, based on the “self -modification” method, resembles a way that people learn to control their facial expressions in the mirror.
The authors of the work emphasize the revolutionary significance of the exact prediction of human facial expressions for the interaction of man and the robot. The integration of human facial expressions in real time not only improves the quality of interaction, but also contributes to the formation of trust between a person and a robot.
Now researchers plan to integrate verbal communication using large language models like ChatGPT in EMO. Researchers emphasize the importance of ethical considerations in the development of technologies that can interpret and imitate human facial expressions with high accuracy, noting that such technologies bring us closer to the future in which robots can naturally integrate into people’s society.