Transfer learning networks with skip connections for classification of brain tumors

This article presents a transfer learning model via convolutional neural networks (CNNs) with skip connection topology, to avoid the vanishing gradient and time complexity, which are usually common in networks. Three pretrained CNN architectures, namely AlexNet, VGG16 GoogLeNet employed equip connections. The is implemented through fine-tuning freezing architectures connections based on magnetic resonance imaging (MRI) slices of brain tumor dataset. Furthermore, preprocessing, frequency-domain information enhancement technique for better image clarity. Performance evaluation conducted obtain improved accuracy MRI classifications.