Implicit Convolutional Kernels for Steerable CNNs
This provides a flexible framework for building general group-equivariant models, benefiting researchers in fields like physics and chemistry, though it is incremental in method.
The authors tackled the problem of implementing steerable convolutional kernels for group-equivariant CNNs, which previously required group-specific analytical solutions, by proposing an implicit neural representation using MLPs that generalizes to any group. They demonstrated effectiveness on tasks like N-body simulations and molecular property prediction, achieving competitive or improved results.
Steerable convolutional neural networks (CNNs) provide a general framework for building neural networks equivariant to translations and transformations of an origin-preserving group $G$, such as reflections and rotations. They rely on standard convolutions with $G$-steerable kernels obtained by analytically solving the group-specific equivariance constraint imposed onto the kernel space. As the solution is tailored to a particular group $G$, implementing a kernel basis does not generalize to other symmetry transformations, complicating the development of general group equivariant models. We propose using implicit neural representation via multi-layer perceptrons (MLPs) to parameterize $G$-steerable kernels. The resulting framework offers a simple and flexible way to implement Steerable CNNs and generalizes to any group $G$ for which a $G$-equivariant MLP can be built. We prove the effectiveness of our method on multiple tasks, including N-body simulations, point cloud classification and molecular property prediction.