Rotation Equivariant Operators for Machine Learning on Scalar and Vector Fields
This work addresses the problem of incorporating rotation symmetry into machine learning models for fields like computer vision and inverse problems, though it is incremental as it builds on existing convolution theorems.
The authors developed theory and software for rotation equivariant operators on scalar and vector fields, enabling spatially invariant dynamics that preserve symmetry, with applications in simulation, optimization, and machine learning, implemented in the Julia package EquivariantOperators.jl for differentiable finite difference operators in 2D/3D.
We develop theory and software for rotation equivariant operators on scalar and vector fields, with diverse applications in simulation, optimization and machine learning. Rotation equivariance (covariance) means all fields in the system rotate together, implying spatially invariant dynamics that preserve symmetry. Extending the convolution theorems of linear time invariant systems, we theorize that linear equivariant operators are characterized by tensor field convolutions using an appropriate product between the input field and a radially symmetric kernel field. Most Green's functions and differential operators are in fact equivariant operators, which can also fit unknown symmetry preserving dynamics by parameterizing the radial function. We implement the Julia package EquivariantOperators.jl for fully differentiable finite difference equivariant operators on scalar, vector and higher order tensor fields in 2d/3d. It can run forwards for simulation or image processing, or be back propagated for computer vision, inverse problems and optimal control. Code at https://aced-differentiate.github.io/EquivariantOperators.jl/