Capsule Networks are this cool new network architecture designed to generalize better than CNNs. For a really good explanation of what's going on please go here: https://www.youtube.com/watch?v=pPN8d0E3900&t=7s
This paper was an investigation in their performance for information retrieval of CAD models. Why that would be useful has to do with how capsule networks handle rotational variance. CNNs basically don't handle it at all.
Although the primary goal of achieving rotation invariance on 3D CAD models wasn't achieved, this project helped me understand better how to approach this kind of thing in the future.
Give it a read and don't be bashful in the comments!