Three dimensional composite images of digital identifications

In general, one innovative aspect of the subject matter described in this specification may be embodied in methods that include generating a three-dimensional composite image of a user from a set of two dimensional facial images. For instance, a depth map may initially be generated for each of the two dimensional facial images based on depth information. The depth maps may be used to identify matching elements that are used to combine multiple two-dimensional images. The generated three-dimensional composite image may then be displayed on a digital identification of a user device. In some instances, the rendering of the three-dimensional composite image on the user device may be adjusted based on tilting motions.