Deep Encoder-Decoder Networks for Mapping Raw Images to Dynamic Movement Primitives

Rok Pahic, Andrej Gams, Ales Ude, Jun Morimoto
2018 2018 IEEE International Conference on Robotics and Automation (ICRA)  
In this paper we propose a new approach for learning perception-action couplings. We show that by collecting a suitable set of raw images and the associated movement trajectories, a deep encoder-decoder network can be trained that takes raw images as input and outputs the corresponding dynamic movement primitives. We propose suitable cost functions for training the network and describe how to calculate their gradients to enable effective training by back-propagation. We tested the proposed
more » ... ach both on a synthetic dataset and on a widely used MNIST database to generate handwriting movements from raw images of digits. The calculated movements were also applied for digit writing with a real robot.
doi:10.1109/icra.2018.8460954 dblp:conf/icra/PahicGUM18 fatcat:s7cefmunbndhvcx4azftcfwgea