Overview / Usage

Convolutional neural networks require a lot of images as training data. Since obtaining such a large amount of
medical data that is labeled by experts is very expensive and difficult, we apply transfer learning to existing
public medical datasets. This research focuses on fine-tuning the latest Imagenet pre-trained model NASNet
by Google followed by a CNN trained medical image data in order to achieve maximum accuracy in
determining the class of the medical image. The expected outcome is to build a model that performs better than existing models in segmenting the medical images. Kindly note that the research is currently being carried upon by us and that we are waiting for the dataset to arrive