Attentive Group Equivariant Convolutional Networks

Abstract

Although group convolutional networks are able to learn powerful representations based on symmetry patterns, they lack explicit means to learn meaningful relationships among them (e.g., relative positions and poses). In this paper, we present attentive group equivariant convolutions, a generalization of the group convolution, in which attention is applied during the course of convolution to accentuate meaningful symmetry combinations and suppress non-plausible, misleading ones. We indicate that prior work on visual attention can be described as special cases of our proposed framework and show empirically that our attentive group equivariant convolutional networks consistently outperform conventional group convolutional networks on benchmark image datasets. Simultaneously, we provide interpretability to the learned concepts through the visualization of equivariant attention maps.

Proceedings of the 37th International Conference on Machine Learning (ICML), 2020

References in corpus (21)

• Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
• Striving for Simplicity: The All Convolutional Net
• Group Equivariant Convolutional Networks
• Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
• On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups
• 3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data
• Rotation equivariant vector field networks
• Stand-Alone Self-Attention in Vision Models
• GCNet: Non-local Networks Meet Squeeze-Excitation Networks and Beyond
• Long Short-Term Memory-Networks for Machine Reading
• Gauge Equivariant Convolutional Networks and the Icosahedral CNN
• Spherical CNNs
• An Empirical Study of Spatial Attention Mechanisms in Deep Networks
• General $E(2)$-Equivariant Steerable CNNs
• Group Equivariant Capsule Networks
• 3D G-CNNs for Pulmonary Nodule Detection
• Deep Scale-spaces: Equivariance Over Scale
• Scale equivariance in CNNs with vector fields
• Co-Attentive Equivariant Neural Networks: Focusing Equivariance On Transformations Co-Occurring In Data
• RotDCF: Decomposition of Convolutional Filters for Rotation-Equivariant Deep Networks
• Affine Self Convolution

Cited by in corpus (8)

• A Data and Compute Efficient Design for Limited-Resources Deep Learning
• Coordinate Independent Convolutional Networks -- Isometry and Gauge Equivariant Convolutions on Riemannian Manifolds
• Alias-Free Convnets: Fractional Shift Invariance via Polynomial Activations
• Group Equivariant Generative Adversarial Networks
• Addressing the Topological Defects of Disentanglement via Distributed Operators
• Group Equivariant Stand-Alone Self-Attention For Vision
• Finite Group Equivariant Neural Networks for Games
• Group Equivariant Subsampling