Abstract | ||
---|---|---|
The use of robots for rehabilitation has become increasingly attractive in recent years. Robots are capable of providing highly repetitive hands-on therapy for patients. In this paper, we present a robotic system for learning a therapist's behavior when interacting with a patient to complete a therapy task. Learning from Demonstration (LfD) techniques are utilized to statistically encode the therapist's behaviors during interaction with a patient. Demonstrations are provided by having the therapist move the patient (and the robot) during the therapy task, which is known as kinesthetic teaching. Later, reproduction of the therapist's interaction is performed by a robot in the absence of the therapist, allowing a patient to continue practicing the therapy task. The results show the system is able to provide interactions similar to the therapist's demonstrated behavior for a given task. |
Year | DOI | Venue |
---|---|---|
2018 | 10.1109/ISMR.2018.8333285 | 2018 International Symposium on Medical Robotics (ISMR) |
Keywords | Field | DocType |
Robotic rehabilitation,kinesthetic teaching,activities of daily living | Kinesthetic learning,Robotic systems,Rehabilitation,Task analysis,Psychology,Learning from demonstration,Rehabilitation robot,Medical treatment,Human–computer interaction,Robot | Conference |
ISBN | Citations | PageRank |
978-1-5386-2513-2 | 1 | 0.39 |
References | Authors | |
7 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jason Fong | 1 | 2 | 2.76 |
Mahdi Tavakoli | 2 | 223 | 49.03 |