Keyframe-based Learning from Demonstration

By Barış Akgün, Maya Cakmak, Karl Jiang, Andrea Lockerd Thomaz. 2012

231 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Control and Systems Engineering, Reinforcement Learning, Robotics & Control

How researchers see this paper

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

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Control and Systems Engineering

In Reinforcement Learning

In Robotics & Control

Citations in the corpus

ROBOTURK: A Crowdsourcing Platform for Robotic Skill Learning through Imitation (2018, background): “Game interfaces frequently introduce artifacts in trajectories that reduce the utility of this data for imitation [36].”
Learning from Physical Human Corrections, One Feature at a Time (2018, background): “Ideally, the human’s correction will only affect the cup’s distance from the table; in practice, however, human actions are noisy and imperfect [7, 19, 20, 24], especially when kinesthetically maneuvering robotic manipulators while trying to carefully orchestrate their multiple degrees of freedom [1”
Deep Imitation Learning for Complex Manipulation Tasks from Virtual Reality Teleoperation (2017, background): “Kinesthetic teaching, in which the human operator guides the robot by force on the robot’s body, can be used to gather demonstrations [9], [11], but is unsuitable for learning visuomotor policies that map from pixels to actions due to the unwanted appearance of human arms.”
Language Conditioned Imitation Learning Over Unstructured Data (2020, background): “Traditionally, imitation learning has been applied to learning individual skills from structured and isolated human demonstrations [39, 1, 53].”
Learning Partially Contracting Dynamical Systems from Demonstrations (2017, methodology): “In [20], a keyframe-based LfD approach is proposed that learns tasks by aligning and clustering the keyframes of obtained demonstrations.”
Neural Task Graphs: Generalizing to Unseen Tasks From a Single Video Demonstration (2018, methodology): “We use N = [1, 5] objects (3-15 steps) out of C = 8 categories, M = N + 3 search locations, and K = 5 drop-off receptacles.”
Learning position and orientation dynamics from demonstrations via contraction analysis (2018, methodology): “Other categories of LfD methods include sampling-based methods (e.g., Bowen and Alterovitz 2014), optimization-based approaches (e.g., Dragan et al. 2015; Zucker et al. 2013), geometric methods (e.g., Ahmadzadeh et al. 2017), and key-frame-based methods (e.g., Akgun et al. 2012).”
Tell me Dave: Context-sensitive grounding of natural language to manipulation instructions (2014, background): “Rather than recording the full movement of the demonstrator which could be not smooth, the recorded sequence for the robot to execute is based on a sequence of keyframe similar to Akgun et al. (2012).”
Helping Robots Learn: A Human-Robot Master-Apprentice Model Using Demonstrations via Virtual Reality Teleoperation (2020, background): “Processing techniques often involve keyframes [21], segmentation [22], [23], constraint extraction [24], [25], and reproduction [26], [27].”
Novel Interaction Strategies for Learning from Teleoperation (2012, methodology): “We presented a method that can learn from hybrid demonstrations in (Akgun et al. 2012a).”
Recent Advances in Robot Learning from Demonstration (2020, background): “As the name would suggest, time-dependent policies rely on time, potentially in addition to any other form of feedback (59, 60, 62, 65).”