Research on Clothing Image Classification by Convolutional Neural Networks

Abstract

In recent years, clothing image classification methods based on convolutional neural networks (CNNs) have attracted plenty of attention with the increasing demand for high accuracy clothing image classification in many fields. In this paper, five different CNNs are designed to implement clothing image classification, which are the conventional CNN, the CNN containing inception module, the CNN containing inception module and residual block, two transfer learned CNNs. The experimental results show that all the networks are capable of achieving good classification, among which the transfer learned CNN have higher classification accuracy.