A Review Study oftheVisual Geometry Group Approachesfor Image Classification
Universiti Tun Hussein Onn Malaysia, Nurzarinah Zakaria, Yana Mazwin Mohmad Hassim, Universiti Tun Hussein Onn Malaysia
Abstract
In the realm of advanced machine learning for image classification, Convolutional Neural Networks (CNNs) stand as a pivotal tool, with the Visual Geometry Group-16(VGG16) model standing out for its emphasis on deepening and expanding CNNsarchitecture to achieve betteraccuracy. However, the complexdesign of VGG16presents challenges regardingcomputational efficiency and scalability. This study addressesthese issues by refining the VGG16architecture through strategic modifications, including reducing convolution blocks, integrating batch normalization (BN) layers, and incorporatinga global average pooling (GAP) layer alongside additional dense and dropout layers.The proposed architecture's effectiveness was assessed through comprehensive experiments across ten benchmark datasets, comparing its performance against the standard VGG16architecture. The proposed architecture sped up the execution time by 63.7% on average across all benchmark datasets, compared to the standard VGG16. Furthermore, the results showed that the proposed architecture outperformed VGG16 by improving the classification accuracy by up to 30.1% based on the overall datasets. In summary, the proposed architectureis made to be compactand accurate. By adjusting parameters, it processes information quickly and accurately. It also includes features to prevent overfitting and improve classification, resulting in a significant advancement in image classification.