Comparative Research on Performance of Image Recognition and Classification Using VGG16 with Different Features

Zhide Ren, Yuxian Wu, Bicheng Zhang

2024

Abstract

Convolutional neural network (CNN) holds a prominent position in machine learning for image recognition and classification. To find out what affects the training efficiency and accuracy, this research primarily enumerates two types of test results based on a typical CNN model called Visual Geometry Group 16 (VGG16) using diverse data sets. At first, VGG16 model is separately introduced towards two data sets. Also, the research presents how VGG16 model works with two data sets. Then comes to the consequences of two data sets, the key points of this comparative research. This research primarily uses accuracy curve and learning curve as evaluation indicators. It not only highlights the challenges that CNN may encounter when dealing with complicated data sets, but also offers a framework for evaluating and comparing the performance of CNN. Throughout this paper, it is concluded that the accuracy rate of different features may present diverse performance. While dealing with data of more weakly correlated features, the VGG16 model may exhibit significantly different accuracy rates in different features compared to other accuracy rates. Furthermore, when compared to data with simple features, the VGG16 model presents lower accuracy in data containing more detailed and complex features.

Paper Citation

in Harvard Style

Ren Z., Wu Y. and Zhang B. (2024). Comparative Research on Performance of Image Recognition and Classification Using VGG16 with Different Features. In Proceedings of the 1st International Conference on Engineering Management, Information Technology and Intelligence - Volume 1: EMITI; ISBN 978-989-758-713-9, SciTePress, pages 95-99. DOI: 10.5220/0012909900004508

in Bibtex Style

@conference{emiti24,
author={Zhide Ren and Yuxian Wu and Bicheng Zhang},
title={Comparative Research on Performance of Image Recognition and Classification Using VGG16 with Different Features},
booktitle={Proceedings of the 1st International Conference on Engineering Management, Information Technology and Intelligence - Volume 1: EMITI},
year={2024},
pages={95-99},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012909900004508},
isbn={978-989-758-713-9},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Engineering Management, Information Technology and Intelligence - Volume 1: EMITI
TI - Comparative Research on Performance of Image Recognition and Classification Using VGG16 with Different Features
SN - 978-989-758-713-9
AU - Ren Z.
AU - Wu Y.
AU - Zhang B.
PY - 2024
SP - 95
EP - 99
DO - 10.5220/0012909900004508
PB - SciTePress