The objective of the project is to increase the cancer detection rate of breast cancer by analyzing CT and biopsied images. This project uses and compares 3 different models to do so.
In this repository, you will find a folder for each model containing the code used to support each of them. The first is leveraging transfer learning from MobileNet. The last layer is trained using the data from this dataset. The rest of the layers are pre-trained. Similarly, the second approach is using transfer learning from Resnet. Again the last layer is trained with the dataset we have collected. Lastly, this project uses a convlutional neural network that has been trained from scratch with only this project's dataset. For this Tensorflow and Keras were used.