Equivariant Representation Learning via Class-Pose Decomposition
This addresses the need for interpretable and disentangled representations in machine learning, though it appears incremental as it builds on existing equivariant learning methods.
The paper tackles the problem of learning equivariant representations for data symmetries by decomposing the latent space into invariant and symmetry components, resulting in representations that outperform other frameworks in capturing data geometry.
We introduce a general method for learning representations that are equivariant to symmetries of data. Our central idea is to decompose the latent space into an invariant factor and the symmetry group itself. The components semantically correspond to intrinsic data classes and poses respectively. The learner is trained on a loss encouraging equivariance based on supervision from relative symmetry information. The approach is motivated by theoretical results from group theory and guarantees representations that are lossless, interpretable and disentangled. We provide an empirical investigation via experiments involving datasets with a variety of symmetries. Results show that our representations capture the geometry of data and outperform other equivariant representation learning frameworks.