We describe and compare several deep-learning approaches for identifying people using 3D point cloud data. We also describe various enhancements that we have made to these networks in order to increase their classification accuracy. Our experiments show that we are able to determine individual identities with high accuracy.