Neuromorphically Inspired Representation for Video Codecs

With recent advancements in Event-vision, the neuromorphic computer vision field which operates with ultra-low electrical requirements, the encoding and decoding of videos as events has become achievable, thus enabling a new form of video codec. This work explores the hypothesis that technologies from multiple field can be combined to make a novel functional AI video codec based on Neuromorphically Inspired Representation (NIR) suited for general use.