3D neuromorphic photonics

Diffraction and scattering are two fundamental physical phenomena used to describe that light will be bent and scattered when it passes through an aperture or encounters an obstacle. By overlapping a multitude of individual and distinct diffractive/scattering signatures, we can create global optical effects as desired. The high number of degrees of freedom and design flexibility in diffractive and scattering elements, open a new avenue in an exciting research area—neuromorphic photonics. In this chapter, we discuss two main categories of 3D neuromorphic computing: diffractive and scattering neural computing. On this foundation, we show the promising applications they enabled, not only in the mainstream inference tasks of recognition, classification, and prediction but also in spatiotemporal wave shaping, optical illusion, logic and mathematic operation, mode detection, and decomposition. Finally, future possible directions are discussed to propel neuromorphic photonics to its full potential.