An Extension of Pre-Trained Deep Learning Model for Sematic Segmentation of Gray-Scale Images 


Vol. 48,  No. 1, pp. 36-39, Jan.  2023
10.7840/kics.2023.48.1.36


PDF
  Abstract

When learning deep learning models on datasets with small amounts of data, utilizing updates and re-learning through transfer learning can reduce learning time and computational resources, and significantly improve the performance of the algorithm. However, most pre-trained models in image processing applications are trained using color images, so three color channels are used as input image values. On the other hand, the gray-scale image is smaller in size than the color image and has one channel, so it cannot be used as an input of a model that has learned several channels. Consequently, existing techniques have limitations in that they need to transform the data to fit the input or modify the layer of the pre-trained model. In this paper, we propose a novel method to use a pre-trained model for semantic segmentation of gray-scale images by adding convolutional layers to the front of a deep learning model. Simulation results show that the proposed scheme has higher accuracy than the existing scheme, and that pre-trained models can be used effectively with good performance.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Related Articles
  Cite this article

[IEEE Style]

J. Jung and Y. Shin, "An Extension of Pre-Trained Deep Learning Model for Sematic Segmentation of Gray-Scale Images," The Journal of Korean Institute of Communications and Information Sciences, vol. 48, no. 1, pp. 36-39, 2023. DOI: 10.7840/kics.2023.48.1.36.

[ACM Style]

Jin-won Jung and Yoan Shin. 2023. An Extension of Pre-Trained Deep Learning Model for Sematic Segmentation of Gray-Scale Images. The Journal of Korean Institute of Communications and Information Sciences, 48, 1, (2023), 36-39. DOI: 10.7840/kics.2023.48.1.36.

[KICS Style]

Jin-won Jung and Yoan Shin, "An Extension of Pre-Trained Deep Learning Model for Sematic Segmentation of Gray-Scale Images," The Journal of Korean Institute of Communications and Information Sciences, vol. 48, no. 1, pp. 36-39, 1. 2023. (https://doi.org/10.7840/kics.2023.48.1.36)
Vol. 48, No. 1 Index

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)






pISSN : 1226 - 4717
eISSN : 2287 - 3880