Retinomorphic Channel Design and Considerations

Abstract

By extrapolating functionality from the retina, retinomorphic engineering has yielded devices that have shown promise to alleviate the challenges presented in modern computer vision tasks. An incredible amount of work has been devoted in recent years to the development and deployment of these event-based vision sensors in applications requiring low-latency, energy-efficient, high dynamic range sensing solutions. However, not much work has been devoted to the area, energy, and speed analysis of the various encoding and decoding mechanisms necessary for sensory pipelines. This paper outlines an empirical framework that presents a clear tradeoff between the various methodologies to transduce physical information in to spikes (encoding) and reconstruct said stimuli from the incident events (decoding). Software-based models of these methodologies were constructed to evaluate the accuracy of stimuli reconstruction for a variety of input profiles. As a result, it is shown that an optimized retinomorphic architecture for a specific set of system-driven cost metrics requires a heterogenous fabric of encoders with a composition of 95% temporal contrast pixels and 5% intensity encoder assuming a temporal jitter of 1μs. Much like the composition of ganglion cells in magno- and parvo-cellular pathway, this multi-modal solution provides the most time, area, and power efficient method to convey visual data.