UCLA researchers have created a hybrid image projection system that combines a neural network encoder with a passive all-optical diffractive decoder to achieve super-resolution images across extended depth of field.