Learning Continuous Human-Robot Interactions from Human-Human Demonstrations
Автор: David Vogt
Загружено: 2015-09-21
Просмотров: 4458
Описание: We present a novel imitation learning approach for learning human-robot interactions from human-human demonstrations. During training, the movements of both interactants are recorded via motion capture and an interaction model is learned that consists of two parts: (1) A Hidden Markov Model enables the robot to identify the relevant interaction example from the pool of all recorded interactions. (2) An interaction mesh for each example that preserves the spatial and temporal details of the shown interaction is used to generate the robot's reaction to the human's motion. The feasibility of the approach is demonstrated with the example of a complex assembly task. We conclude that our approach is well suited for collaborative tasks requiring continuous body movement coordination of human and robot.
Повторяем попытку...
Доступные форматы для скачивания:
Скачать видео
-
Информация по загрузке:
