DeepSphere: a graph-based spherical CNN

By Michaël Defferrard, Martino Milani, Frédérick Gusset, Nathanael Perraudin. 2020

103 citations recorded by the source corpus.

Designing a convolution for a spherical neural network requires a delicate tradeoff between efficiency and rotation equivariance. DeepSphere, a method based on a graph representation of the discretized sphere, strikes a controllable balance between these two desiderata. This contribution is twofold. First, we study both theoretically and empirically how equivariance is affected by the underlying graph with respect to the number of pixels and neighbors. Second, we evaluate DeepSphere on relevant problems. Experiments show state-of-the-art performance and demonstrates the efficiency and flexibility of this formulation. Perhaps surprisingly, comparison with previous work suggests that anisotropic filters might be an unnecessary price to pay.

Subfields: Computational Mechanics, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

1. Bo Yang
2. Yulan Guo
3. Andrew Markham
4. Sheng Ao
5. Qingyong Hu

In Computer Vision and Pattern Recognition

1. Bo Yang
2. Yulan Guo
3. Andrew Markham
4. Sheng Ao
5. Qingyong Hu

Citations in the corpus

• Group Equivariant Stand-Alone Self-Attention For Vision (2020, neutral): “, point clouds (Thomas et al., 2018; Defferrard et al., 2020; Finzi et al., 2020; Fuchs et al., 2020).”
• SpinNet: Learning a General Surface Descriptor for 3D Point Cloud Registration (2020, background): “Accurate matching of partial 3D surfaces is critical for point cloud registration [17, 38, 6, 21, 29], segmentation [58, 28, 27], and recognition [24, 12, 47].”