BROWSE

ITEM VIEW & DOWNLOAD

A convolutional neural network based policy inspired by the cerebellum

Cited 0 times inthomson ciCited 0 times inthomson ci
Title
A convolutional neural network based policy inspired by the cerebellum
Other Titles
소뇌 구조 모방 합성곱 신경망 기반의 제어 정책
Author
Shin, Dong-Ju
Advisor
Kim, Kwang-In
Issue Date
2020-08
Publisher
Graduate School of UNIST
Abstract
With the recent development of deep neural networks, the demands and expectations of using deep learning in robotic behavior learning are increasing. Methods for controlling robots based on visual information using deep learning-based models such as deep visuomotor policy have been actively explored. However, these methods do not consider much about a good neural network structure to process robot configuration input. In this paper, we present a novel neural network inspired by human cerebellum which contains 1-dimensional convolution layers. We evaluate our model on various real-world and simulation tasks. It is experimentally demonstrated that our novel neural network has better expressiveness for joint information than previous models.
Description
Department of Computer Science and Engineering
URI
Go to Link
Appears in Collections:
CSE_Theses_Master
Files in This Item:
A convolutional neural network based policy inspired by the cerebellum.pdf Download

find_unist can give you direct access to the published full text of this article. (UNISTARs only)

Show full item record

qrcode

  • mendeley

    citeulike

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

MENU