Imaging applications through learning-enhanced metaoptics

Metaoptics can provide outstanding solutions for imaging applications in extended depth of field, light field microscopy, and hyperspectral imaging. A series of conventional optical components, such as lenses and mirrors, are employed in traditional optics solutions. In pursuit of a more compact form factor, we demonstrate that the metasurface optics provide key advantages in imaging, such as extreme Extended Depth of Field (EDOF), where the extended DOF range is well beyond what is demonstrated in state-of-the-art. This can be further implemented in a light field microscope to obtain a compact light field microscopy with high-speed volumetric imaging and high spatio-temporal resolution. Finally, I will demonstrate how metaoptic components can be utilized for a compact and efficient snapshot hyperspectral camera model. In all of these examples, we take advantage of end-to-end design. We have developed a fully-differentiable wave optics-based deep learning framework, combining novel hardware (metaoptics) and software (artificial neural computing).