Imitation learning for coordinated human–robot collaboration based on hidden state-space models

Wang, Likun; Wang, Guoyan; Jia, Shuya; Turner, Alison; Ratchev, Svetan

doi:10.1016/j.rcim.2021.102310

Cited by 8 publications

(1 citation statement)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In many human-robot interaction scenarios, robots need to be able to imitate human motions [3,20]. For example, imitating human motions can be used to reproduce human demonstrated motions [9,23], for coordination purposes in human-robot collaboration [26], or even to provide feedback to patients in physical rehabilitation scenarios [4]. This imitation is not a simple one-to-one mapping from human joint angle to motor angle, as the embodiment of humans and robots differ in sizes, proportions, velocities, forces, dynamics.…”

Section: Introductionmentioning

confidence: 99%

Unsupervised Motion Retargeting for Human-Robot Imitation

Annabi,

Ma,

Nguyen

2024

Companion of the 2024 ACM/IEEE International Conference on Human-Robot Interaction

View full text Add to dashboard Cite

This early-stage research work aims to improve online humanrobot imitation by translating sequences of joint positions from the domain of human motions to a domain of motions achievable by a given robot, thus constrained by its embodiment. Leveraging the generalization capabilities of deep learning methods, we address this problem by proposing an encoder-decoder neural network model performing domain-to-domain translation. In order to train such a model, one could use pairs of associated robot and human motions. Though, such paired data is extremely rare in practice, and tedious to collect. Therefore, we turn towards deep learning methods for unpaired domain-to-domain translation, that we adapt in order to perform human-robot imitation. CCS CONCEPTS• Computer systems organization → Robotics.

show abstract