Recognizing teleoperated manipulations

By P. Pook, Dana H. Ballard. 1993

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Visual Imitation Made Easy (2020, methodology): “This method has been successfully applied to a large range of robotic tasks including ﬂying a robotic helicopter [17], grasping objects [18, 19], navigating robots through cluttered environments [20, 21, 22], and even driving cars [23].”
A survey of robot learning from demonstration (2009, unknown): “…via a joystick are used in a variety of applications, including flying a robotic helicopter [6], robot kicking motions [7], object grasping [8] [9], robotic arm assembly tasks [10] and obstacle avoidance and navigation [11] [12] Categorization of approaches to building the demonstration…”
Machine learning of motor skills for robotics (2007, methodology): “Pook and Ballard (1993) created a robotic system that initially discretized demonstrated movement and force trajectories by means of vector quantization, and then trained a hidden Markov model (HMM) with these discrete states as observables to find the transition probabilities among a given set of m”
A survey of Tactile Human-Robot Interactions (2010, unknown): “demonstrations [64,65].”
Embodied Symbol Emergence Based on Mimesis Theory (2004, methodology): “Several researches have already proposed motion recognition methods based on the HMM (Yamato, Ohya, and Ishii 1992; Pook and Ballard 1993; Wada and Matsuyama 1998; Yoshiike et al. 1998; Ogawara et al. 2002), however, no research exists in which motion is generated from the HMM.”
Imitation Learning (2017, methodology): “Imitation learning research done in this direction focuses on object manipulation, such as sorting and assembly tasks [Saunders et al. 2006; Pook and Ballard 1993; Oztop and Arbib 2002; Calinon and Billard 2007b; Nicolescu and Mataric 2003].”
Behavior recognition for humanoid robots using long short-term memory (2016, background): “A number of researches agreed that one common way to overcome this is to transform the demonstration into a set of preprogrammed higher-level actions called subtasks, motion primitives, motor primitives or motor skills.(2,3) This transformation supports not only generalization purposes but also as a”