Mask R-CNN builds upon the massively popular Faster R-CNN model, which was not designed for “pixel-to-pixel alignment between network inputs and outputs”, by adding a mask prediction branch for ...

As the title suggests, this paper studies various different deep convolutional neural network architectures and various techniques to use these CNNs for CADe (Computer Aided Detection) tasks. With the ...