Knee Images Classification Using Transfer Learning

EasyChair Preprint 4588

Abstract

Deep learning techniques are very common and mostly used in the last few years because it has a good ability for feature extraction. Convolution Neural Network (CNN) is a deep learning technique, which concerns with classification problems. In this work, the transfer learning technique is applied in training a CNN to classify images of knee to their respective classes. The proposed method uses the NASNet Mobile Network pre-trained Convolution Neural Network.

Keyphrases: Convolutional Neural Network, Transfer Learning, machine learning, medical image classification

@booklet{EasyChair:4588,
  author    = {Kamel Rahouma and Ahmed Salama},
  title     = {Knee Images Classification Using Transfer Learning},
  howpublished = {EasyChair Preprint 4588},
  year      = {EasyChair, 2020}}