Trained VGG19 model for multiclass histology image classification (Matlab)
Jakob Nikolas Kather · Zenodo (CERN European Organization for Nuclear Research) · 2018
This is a convolutional neural network model for Matlab. It can be used to classify histological image patches of human colorectal cancer. To create this, we have used a VGG19 model from the Matlab R2017b (Mathworks, Natick, MA, USA) Add-On Explorer and re-trained it with the normalized NCT-100K data set (DOI 10.5281/zenodo.1214456). Please observe that the original model was shared under the "CC BY 4.0" license. We hereby give credit to the original researchers: http://www.robots.ox.ac.uk/~vgg/research/very_deep