Tuning Modular Networks With Weighted Losses for Hand-Eye Coordination

Fangyi Zhang, Jurgen Leitner, Michael Milford, Peter I. Corke; Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2017, pp. 24-25

Abstract


This paper introduces an end-to-end fine-tuning method to improve hand-eye coordination in modular deep visuo-motor policies (modular networks) where each module is trained independently. Benefiting from weighted losses, the fine-tuning method significantly improves the performance of the policies for a robotic planar reaching task.

Related Material


[pdf] [arXiv]
[bibtex]
@InProceedings{Zhang_2017_CVPR_Workshops,
author = {Zhang, Fangyi and Leitner, Jurgen and Milford, Michael and Corke, Peter I.},
title = {Tuning Modular Networks With Weighted Losses for Hand-Eye Coordination},
booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
month = {July},
year = {2017}
}