Learning in Robot Vision Directed Reaching: A Comparison of Methods
Abstract
Four neural network algorithms were examined for their ability to adaptively associate stereo camera coordinates with joint positions of a three degree of freedom manipulator arm in a 3D reaching task. Given reasonable numbers of training exemplars for an implementation in real hardware, all networks trained to significant errors. Two secondary error correction procedures were then tested. Both further reduced errors, but one method that depended on continuous visual and proprioceptive feedback to train a small set of associative weights that correlated joint and camera velocities was especially effective in eliminating errors. Stereo pan, tilt and vergence information was used to direct ballistic reaching, but relative depth information, was used for the visual feedback of end-effector velocity in the second error correction method.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 1994
- Accession Number
- ADA483612
Entities
People
- Hoa G. Nguyen
- Michael R. Blackburn
Organizations
- Naval Command, Control and Ocean Surveillance Center