Home
Author Guide
Editor Guide
Reviewer Guide
Published Issues
Special Issue
Introduction
Special Issues List
Sections and Topics
Sections
Topics
journal menu
Aims and Scope
Editorial Board
Indexing Service
Article Processing Charge
Open Access
Copyright and Licensing
Preservation and Repository Policy
Publication Ethics
Editorial Process
Contact Us
General Information
ISSN:
1798-2340 (Online)
Frequency:
Monthly
DOI:
10.12720/jait
Indexing:
ESCI (Web of Science)
,
Scopus
,
CNKI
,
etc
.
Acceptance Rate:
19%
APC:
500 USD
Average Days to Accept:
135 days
Journal Metrics:
Impact Factor 2022: 1.0
3.1
2022
CiteScore
49th percentile
Powered by
Editor-in-Chief
Prof. Kin C. Yow
University of Regina, Saskatchewan, Canada
I'm delighted to serve as the Editor-in-Chief of
Journal of Advances in Information Technology
.
JAIT
is intended to reflect new directions of research and report latest advances in information technology. I will do my best to increase the prestige of the journal.
What's New
2024-03-28
Vol. 15, No. 3 has been published online!
2024-02-26
The papers published in Vol. 15, Nos. 1&2 have been registered with Crossref.
2024-02-26
Vol. 15, No. 2 has been published online!
Home
>
Published Issues
>
2020
>
Volume 11, No. 4, November 2020
>
Improving the Representation of CNN Based Features by Autoencoder for a Task of Construction Material Image Classification
S. Bunrit, N. Kerdprasop, and K. Kerdprasop
Suranaree University of Technology, Thailand
Abstract
—Deep learning based model named Convolution Neural Network (CNN) has been extensively employed by diversified applications concerned images or videos data. Because training a specific CNN model for an application task consumes enormous machine resources and need many of the training data, consequently pre-trained models of CNN have been broadly used as the transfer-learning scenario. By the scenario, features had been learned from a pre-trained model by one source task can be proficiently sent further to another specific task in a concept of knowledge transferring. As a result, a task specific can be directly employed such pre-trained features or further train more by setting the pre-trained features as a starting point. Thereby, it takes not much time and can improve the performance from many referenced works. In this work, with a task specific on construction material images classification, we investigate on the transfer learning of GoogleNet and ResNet101 that pre-trained on ImageNet dataset (source task). By applying both of the transfer-learning schemes, they reveal quite satisfied results. The best for GoogleNet, it gets 95.50 percent of the classification accuracy by fine-tuning scheme. Where, for ResNet101, the best is of 95.00 percent by using fixed feature extractor scheme. Nevertheless, after the learning based representation methods are further employed on top of the transferred features, they expose more appeal results. By Autoencoder based representation method reveals the performance can improve more than PCA (Principal Component Analysis) in all cases. Especially, when the fixed feature extractor of ResNet101 is used as the input to Autoencoder, the classified result can be improved up to 97.83%. It can be inferred, just applying Autoencoder on top of the pre-trained transferred features, the performance can be improved by we have no need to fine-tune the complex pre-trained model.
Index Terms
—Convolution Neural Network (CNN), transfer learning, Autoencoder, construction material, image classification
Cite: S. Bunrit, N. Kerdprasop, and K. Kerdprasop, "Improving the Representation of CNN Based Features by Autoencoder for a Task of Construction Material Image Classification," Journal of Advances in Information Technology, Vol. 11, No. 4, pp. 192-199, November 2020. doi: 10.12720/jait.11.4.192-199
Copyright © 2020 by the authors. This is an open access article distributed under the Creative Commons Attribution License (
CC BY-NC-ND 4.0
), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.
2-CE030_Thailand
PREVIOUS PAPER
Blockchain-Based Access Control Supporting Anonymity and Accountability
NEXT PAPER
Determining Smokers’ Quitting Behavior Patterns for Multi-level Intervention of a Smoking Cessation Program