WDN: A Wide and Deep Network to Divide-and-Conquer Image Super-Resolution

Divide and conquer is an established algorithm design paradigm that has proven itself to solve a variety of problems efficiently. However, it yet be fully explored in solving with neural network, particularly the problem image super-resolution. In this work, we propose approach divide super-resolution into multiple subproblems then solve/conquer them help network. Unlike typical deep alternate network architecture much wider (along being deeper) than existing networks specially designed implement divide-and-conquer Additionally, technique calibrate intensities feature map pixels introduced. Extensive experimentation on five datasets reveals our towards proposed generate better sharper results current state-of-the-art methods.