The Animal Classification: An Evaluation of Different Transfer Learning Pipeline

Authors

  • Ken-ji Ee Faculty of Manufacturing and Mechatronic Engineering Technology, Universiti Malaysia Pahang, 26600 Pekan Pahang, Malaysia.
  • Ahmad Fakhri Bin Ab. Nasir Innovative Manufacturing, Mechatronics and Sport Laboratory, Faculty of Manufacturing Engineering, Universiti Malaysia Pahang, Pekan, Malaysia.
  • Anwar P. P. Abdul Majeed Faculty of Manufacturing and Mechatronic Engineering Technology, Universiti Malaysia Pahang, 26600 Pekan Pahang, Malaysia.
  • Mohd Azraai Mohd Razman Faculty of Manufacturing and Mechatronic Engineering Technology, Universiti Malaysia Pahang, 26600 Pekan Pahang, Malaysia.
  • Nur Hafieza Ismail Centre for Software Development & Integrated Computing, Universiti Malaysia Pahang, 26600 Malaysia.

DOI:

https://doi.org/10.15282/mekatronika.v3i1.6680

Keywords:

animal classification, transfer learning, VGG model, SVM, k-NN

Abstract

The animal classification system is a technology to classify the animal class (type) automatically and useful in many applications. There are many types of learning models applied to this technology recently. Nonetheless, it is worth noting that the extraction of the features and the classification of the animal features is non-trivial, particularly in the deep learning approach for a successful animal classification system. The use of Transfer Learning (TL) has been demonstrated to be a powerful tool in the extraction of essential features. However, the employment of such a method towards animal classification applications are somewhat limited. The present study aims to determine a suitable TL-conventional classifier pipeline for animal classification. The VGG16 and VGG19 were used in extracting features and then coupled with either k-Nearest Neighbour (k-NN) or Support Vector Machine (SVM) classifier. Prior to that, a total of 4000 images were gathered consisting of a total of five classes which are cows, goats, buffalos, dogs, and cats. The data was split into the ratio of 80:20 for train and test. The classifiers hyper parameters are tuned by the Grids Search approach that utilises the five-fold cross-validation technique. It was demonstrated from the study that the best TL pipeline identified is the VGG16 along with an optimised SVM, as it was able to yield an average classification accuracy of 0.975. The findings of the present investigation could facilitate animal classification application, i.e. for monitoring animals in wildlife.

Downloads

Published

2021-06-17

How to Cite

[1]
K.- ji Ee, A. F. B. Ab. Nasir, A. P. P. Abdul Majeed, M. A. Mohd Razman, and N. H. . Ismail, “The Animal Classification: An Evaluation of Different Transfer Learning Pipeline”, MEKATRONIKA, vol. 3, no. 1, pp. 27–31, Jun. 2021.

Issue

Section

Original Article

Most read articles by the same author(s)

<< < 1 2 3 > >>