An intuitive multi-frequency feature representation for SO(3)-equivariant networks

arXiv:2405.04537v1 Announce Type: new
Abstract: The usage of 3D vision algorithms, such as shape reconstruction, remains limited because they require inputs to be at a fixed canonical rotation. Recently, a simple equivariant network, Vector Neuron (VN) has been proposed that can be easily used with the state-of-the-art 3D neural network (NN) architectures. However, its performance is limited because it is designed to use only three-dimensional features, which is insufficient to capture the details present in 3D data. In this paper, …

abstract algorithms architectures art arxiv canonical cs.ai cs.cv cs.gr feature inputs network networks neural network neuron representation rotation simple state type usage vector vision