基于VGG-16的电商评论图像审核

E-commerce review image filtering based on Transfer Learning

如今互联网电商平台中用户对于购买商品上传的点评图片质量参差不齐,影响其他用户的购物体验和对于商品质量的判断,电商公司通常通过人工审核来规避这种情况,然而大量的上传图片数据需要大量的人力进行运营审核,针对平台当前审核成本过高的问题,本文设计了一种基于VGG-16卷积神经网络的电商评论图像分类模型,并采用随机梯度下降算法、防止过拟合技术来改进模型,通过迁移学习方法对评论图片进行识别分类从而实现评论图像的自动审核。实验结果显示,本研究模型相比其他传统网络模型效果更好,具有很高的识别精度、鲁棒性和泛化能力,可以准确快速完成对评论图像的分类筛选,且具有一定的扩展性。

In order to solve the problem that the quality of the comments and pictures uploaded by users on the Internet e-commerce platform is uneven,which affects the shopping experience of other users and the judgment of product quality,this paper designs an e-commerce review image classification model based on vgg-16 convolutional neural network,and uses the random gradient descent algorithm to improve the model and prevent over fitting technology Mobile learning method is used to recognize the comment images.The experimental results show that the proposed model is better than other traditional network models,and has high recognition accuracy,robustness and generalization ability.It can accurately and quickly complete the classification and screening of review images,and has certain scalability.