Retinal and Neural Dynamics using Memristor-CMOS Architectures

This thesis is dedicated to the application of ReRAM and memristive systems, to reconstructing the neural dynamics within the retina and the brain. In its present operative form, visual perception by way of prosthesis is limited to flashes of light and shapes of varying sizes, locations, and brightness. By developing a neuromorphic vision chip with memristor-CMOS hybrid technology that gives a spike-based response as inspired by the ganglion cells, we are able to converge towards designing smart image sensors that operate at high resolutions with ultra-low power, much like the retina.