3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data - 42Papers