FIG. 3 illustrates an example of point cloud alignment with basis set learning according to embodiments of the present disclosure. As described above with regard to FIG. 1, in some examples the system 100 may generate output data such as mesh data 118, which may correspond to an output mesh, 3D scan, 3D model, deformable model, reposable human avatar, and/or the like. The mesh data 118 may include detailed geometry and/or color information representing an appearance of the user 5 and may be used to represent the user 5 using 3D surfaces.

As illustrated in FIG. 3, the system 100 may process an input point cloud 310 using a fixed set of basis points 320 and generate distance values 330. For example, after receiving permission from the user 5, the system 100 may generate the input point cloud 310 and then process the input point cloud 310 to generate the distance values 330. The system 100 may input the distance values 330 to a deep neural network (DNN) encoder 340, which may process the distance values 330 and generate mesh vertices 350 (e.g., the mesh data 118).