High-density salt & pepper noise removal using machine learning

This paper proposes an additional ‘Median Layer’ to the existing Convolutional Neural Network (CNN).The work primarily focuses on removing ‘High Density Impulse Noise’ present in image. The proposed median layer performs filtering all feature channels. developed network is end-to-end that does not require any non-trivial preprocessing tasks. Based experiments performed, results show that, by inserting a CNN, architecture has given convincing results, when compared approaches.