Performance Analysis of CNN Channel Attention Modules for Image Classification Task

Authors : Mehmet SARIGÜL

Pages : 35-40

Doi:10.21605/cukurovaumfd.1273688

View : 10 | Download : 6

Publication Date : 2023-03-30

Article Type : Research Paper

Abstract :Increasing the representation power of convolutional neural networks, a popular deep learning model, is one of the hot study topics recently. Channel attention is a common strategy followed in this regard. In this strategy, the inter-channel relationship is exploited by a module placed after the convolution operation. Recently, successful channel attention modules are proposed in this context. In this article, a performance analysis of three popular channel attention structures which are Squeeze-and-Excitation Networks insert ignore into journalissuearticles values(SeNet);, Efficient Channel Attention Networks insert ignore into journalissuearticles values(Eca-Net);, and Convolutional Block Attention Module insert ignore into journalissuearticles values(CBAM);, is performed using five different image datasets for the classification task. According to the obtained results, SeNet is the most successful channel attention module surpassing the other’s performance in the majority of the experiments. In experiments with the ResNet18 and ResNet34 base models, the SeNet module showed the highest performance in three of the five datasets. For the ResNet50 baseline, SeNet was the channel attention module with the highest accuracy values for all datasets.
Keywords : Evrişimli sinir ağları, Kanal dikkati, Görüntü sınıflandırması