Research on E-Commerce Platform Product Image Recognition Based on ResNet Network

Kunxian Wang

2023

Abstract

With the development and popularity of the Internet and smart mobile devices, online shopping has gradually become the main way for contemporary people to shop, the massive amount of commodity information makes it difficult for users to choose, how to get the correct commodity information has become a difficult problem, and manual labelling of commodity categories is very inefficient. How to improve the correct rate of commodity classification, numerous academics have conducted extensive research in this field., the current mainstream method is to use convolutional neural networks for commodity picture classification. In this paper, we carry out a study on the classification of commodity pictures on e-commerce platforms by constructing three ResNet networks with different depths, ResNet18, ResNet34 and ResNet50, respectively, and explore the practical significance by training and comparing the classification accuracy by using the commodity pictures directly downloaded from the famous e-commerce platforms in China. From the experimental results, evidently, as the depth of network layers increases, the performance results of the network are getting better and better, while all three networks have achieved a high accuracy rate, which indicates that convolutional neural networks have application value in the classification of commodity pictures.

Paper Citation

in Harvard Style

Wang K. (2023). Research on E-Commerce Platform Product Image Recognition Based on ResNet Network. In Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML; ISBN 978-989-758-705-4, SciTePress, pages 87-91. DOI: 10.5220/0012804000003885

in Bibtex Style

@conference{daml23,
author={Kunxian Wang},
title={Research on E-Commerce Platform Product Image Recognition Based on ResNet Network},
booktitle={Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML},
year={2023},
pages={87-91},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012804000003885},
isbn={978-989-758-705-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML
TI - Research on E-Commerce Platform Product Image Recognition Based on ResNet Network
SN - 978-989-758-705-4
AU - Wang K.
PY - 2023
SP - 87
EP - 91
DO - 10.5220/0012804000003885
PB - SciTePress