Computer Vision Fall 2018 Project

Authors: Yair Schiff ([email protected]) and Joanna Bitton ([email protected])

Instructor: Prof. Rob Fergus

Overview:

This repository applies the Mask R-CNN architecture and other Convolutional Neural Network architectures to breast mammography data, namely the "Digital Database for Screening Mammography" (DDSM) and INbreast datasets.

• The data preparation directory contains scripts for processing the DDSM and INbreast data set for the segmentation and classification tasks.
• The classification directory contains scripts for training and evaluating classification of mass presence in a mammography image.
• The Mask R-CNN directory contains scripts for running the matterport implemenation of the Mask R-CNN algorithm on the mammography data.