Semantic segmentation is one of the major problems in the computer vision domain. Solving this task effectively is a step towards more complex tasks such as scene understanding and object interactions recognition from visual data. Semantic segmentation is applicable in many important real world applications as well. Self-driving vehicles and robot vision, which can have great impact on the living quality, is among those fields that can benefit from semantic segmentation. On the other hand, advances in deep learning have caused noticeable improvements in many AI subfields such as computer vision. Semantic segmentation has also enjoyed the recent advances of deep neural networks, fully convolutional networks, and new methods based on deep learning, which has surpassed their Segmentation, Fully Convolutional Networks, Memory Augmented Neural Network Key Words : Semantic Segmentation, Fully Convolutional Networks, Memory Augmented Neural Networks
