ameya005/Deep-Segmentation

Deep Learning methods for semantic segmentation with weakly labelled data

Deep-Segmentation

Deep Learning methods for semantic segmentation with weakly labelled data

Implementation of

Pinheiro, P. and Collobert, R.; From Image-level to Pixel-level Labeling with Convolutional Networks

The paper uses OverFeat. This implementation tests the algorithm out with the cifar10_cnn example network from keras. The modifications are applied on top of the same.