The article focuses on the use of convolutional neural networks for matching product images. The importance of developing systems for products image matching is described. The analysis of image comparison methods is carried out, the advantages and disadvantages of each method are noted. The comparison of matching products images from retailers using the ResNet neural network has been executed. Experiments and testing on datasets of products from retailers for image matching have been carried out with ResNet neural networks, the accuracy of image matching for different architectures of the ResNet neural network has been examined. Conclusions about the possibility of using the ResNet neural networks for matching product images were made.
Keywords: convolutional neural network, recognition, image, image analysis, image matching, ResNet