Principal Warps: Thin-Plate Splines and the Decomposition of Deformations
IEEE Transactions on Pattern Analysis and Machine Intelligence
Explanation-Based Generalization: A Unifying View
Machine Learning
Interactive Robot Programming Based on Human Demonstration and Advice
Selected Papers from the International Workshop on Sensor Based Intelligent Robots
Imitation in animals and artifacts
A Four-step Camera Calibration Procedure with Implicit Image Correction
CVPR '97 Proceedings of the 1997 Conference on Computer Vision and Pattern Recognition (CVPR '97)
Object Recognition from Local Scale-Invariant Features
ICCV '99 Proceedings of the International Conference on Computer Vision-Volume 2 - Volume 2
Natural methods for robot task learning: instructive demonstrations, generalization and practice
AAMAS '03 Proceedings of the second international joint conference on Autonomous agents and multiagent systems
A new point matching algorithm for non-rigid registration
Computer Vision and Image Understanding - Special issue on nonrigid image registration
Real-Time Non-Rigid Surface Detection
CVPR '05 Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Volume 1 - Volume 01
Neural Networks - 2006 Special issue: The brain mechanisms of imitation learning
Multi-robot plan adaptation by constrained minimal distortion feature mapping
ICRA'09 Proceedings of the 2009 IEEE international conference on Robotics and Automation
Pattern Recognition
Hi-index | 0.00 |
We propose a novel approach for robot movement imitation that is suitable for robotic arm movement in tasks such as reaching and grasping. This algorithm selects a previously observed path demonstrated by an agent and generates a path in a novel situation based on pairwise mapping of invariant feature locations present in both the demonstrated and the new scenes using minimum distortion and minimum energy strategies. This One-Shot Learning algorithm is capable of not only mapping simple point-to-point paths but also adapting to more complex tasks such as involvement of forced waypoints. As compared to traditional methodologies, our work does not require extensive training for generalisation as well as expensive run-time computation for accuracy. Cross-validation statistics of grasping experiments show great similarity between the paths produced by human subjects and the proposed algorithm.