The Effect of Filter Usage on the Results in Image Classification with Artificial Intelligence/Deep Learning Methods

Oncevarlik A., Akon A., Adali E.

7th International Conference on Computer Science and Engineering, UBMK 2022, Diyarbakır, Turkey, 14 - 16 September 2022, pp.450-455 identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/ubmk55850.2022.9919549
  • City: Diyarbakır
  • Country: Turkey
  • Page Numbers: pp.450-455
  • Keywords: AI, AlexNet, Deep Learning, Effect of filters, Image Classification
  • Istanbul Technical University Affiliated: Yes

Abstract

© 2022 IEEE.Various methods are used to classify documents. Some of the methods are image-based and some are text-based. In this study, it is introduced how image methods can be used in the classification of documents. The results were evaluated by using artificial intelligence, deep learning methods and different filters, especially used in image classification studies. In this study, images are classified based on the implementation of the AlexNet CNN algorithm on Tensorflow. At this stage, the effect on the results was measured by using different filters. In this paper, instead of introducing a new method, the effects and results of various filters applied on the known AlexNet CNN model are introduced. Thus, the comparison of filters is included.