The evolution of modern-day science has come a long way to understanding several problems, but we still lack a detailed understanding of the human visual system. One of the plausible directions of study concentrates on identifying the incompleteness of the vision system by tricking the retinal input to create illusory experiences. One such visual illusion - Hermann grid - is yet to be understood properly. Baumgartner’s retinal localization theory lacks a satisfactory explanation. Several critical studies hinted towards the importance of edges and corners (edge of the edges) for the replication or reduction of illusory experiences. One of the low-level visual tasks is deblurring, which helps in edge enhancement and image sharpening, and we conjectured that this functionality might engender or lessen the illusory experiences. We implemented a deep learning-based computational model – deblurring U-Net deep learning network which substantiates our claim by replicating perceptually similar experiences regarding the grid illusion and its variants. In the future, we aim to explain the colored grid illusions and try to peep into the individual layers of the U-Net to analyze the feature maps.