@inproceedings{Wu:2010:10.1109/ICARCV.2010.5707349,
author = {Wu, Y and Demiris, Y},
doi = {10.1109/ICARCV.2010.5707349},
pages = {453--458},
publisher = {IEEE},
title = {Hierarchical Learning Approach for One-shot Action Imitation in Humanoid Robots},
url = {http://dx.doi.org/10.1109/ICARCV.2010.5707349},
year = {2010}
}

Download

TY  - CPAPER
AB  - We consider the issue of segmenting an action in the learning phase into a logical set of smaller primitives in order to construct a generative model for imitation learning using a hierarchical approach. Our proposed framework, addressing the “how-to” question in imitation, is based on a one-shot imitation learning algorithm. It incorporates segmentation of a demonstrated template into a series of subactions and takes a hierarchical approach to generate the task action by using a finite state machine in a generative way. Two sets of experiments have been conducted to evaluate the performance of the framework, both statistically and in practice, through playing a tic-tac-toe game. The experiments demonstrate that the proposed framework can effectively improve the performance of the one-shot learning algorithm and reduce the size of primitive space, without compromising the learning quality.
AU  - Wu,Y
AU  - Demiris,Y
DO  - 10.1109/ICARCV.2010.5707349
EP  - 458
PB  - IEEE
PY  - 2010///
SP  - 453
TI  - Hierarchical Learning Approach for One-shot Action Imitation in Humanoid Robots
UR  - http://dx.doi.org/10.1109/ICARCV.2010.5707349
UR  - http://hdl.handle.net/10044/1/12666
ER  - 

Download