Deep Convolutional Encoder-Decoder Architecture implemented along with max-pooling indices for pixel-wise semantic segmentation using CamVid dataset.