발간년도 : [2023]
| 논문정보 |
|
| 논문명(한글) |
[Vol.18, No.6] A Study on the Transfer Learning-based Humanoid Robot HRI-Vision System |
|
| 논문투고자 |
Won Kim |
|
| 논문내용 |
With the development of artificial intelligence many changes are taking place in various fields andservice robots equipped with machine learning and industrial cooperative robots are also appearing in therobot field. However the intuitive and empathic intelligence of artificial intelligence has not yet fullypermeated service robots and surpassed humans. Therefore it can be said that it is difficult for robots toperform complex services independently and robots and humans must reinforce each other to providequality services to customers. Recently HRI a convergence technology for natural communicationbetween humans and robots and smooth interaction are attracting attention. HRI research can be dividedinto user-centered and robot-centered interactions. The robot-vision technology recognizes the environmentand objects, recognizes a human face to identify a target, recognizes emotions and plays the role of arecognition unit that can infer intentions based on motion. Robots are now so widespread that humansconsider them to be their colleagues and friends. Therefore, there is a need for a small robotincorporating object recognition technology. In this paper, we propose an HRI vision object recognitionsystem using a transfer learning-based object detection algorithm in a small PC of a humanoid robot.The proposed system applies robot vision technology that can replace the visual functions of low-specsmall humanoid robots and allows the humanoid robot to perform pre-programmed motions through facerecognition. |
|
| 첨부논문 |
|
|
|
|
|
자료실 > 논문지 
