NAO and Expert Imitating Each Other: A New Robot Action Dataset
[ X ]
Tarih
2025
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Institute of Electrical and Electronics Engineers Inc.
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
Today, with the rapid spread of robots, the significance of robot training has increased. Imitation learning plays a significant role in robot training, and numerous scientific studies are published in this field each year. In this method, robots can be controlled by imitating human movements and this process is called teleoperation. In teleoperation studies, the mapping between human and robot skeletal systems is performed using two fundamental approaches: the mathematical model and the artificial intelligence model. However, the number of datasets in the literature that simultaneously record human and robot skeletal data to enable a comprehensive comparison of these two approaches is quite limited. In this study, we created two separate datasets containing human and robot skeletal data and experimentally compared two different teleoperation methods. The experimental findings indicate that the artificial intelligence model, when trained with unbiased datasets, outperforms the mathematical model in teleoperation processes. © 2025 IEEE.
Açıklama
33rd IEEE Conference on Signal Processing and Communications Applications, SIU 2025 -- 2025-06-25 through 2025-06-28 -- Istanbul -- 211450
Anahtar Kelimeler
artificial intelligence, dataset, humanoid robot, teleoperation
Kaynak
WoS Q Değeri
Scopus Q Değeri
N/A
