RotDCF: Decomposition of Convolutional Filters for Rotation-Equivariant Deep Networks

Abstract

Explicit encoding of group actions in deep features makes it possible for convolutional neural networks (CNNs) to handle global deformations of images, which is critical to success in many vision tasks. This paper proposes to decompose the convolutional filters over joint steerable bases across the space and the group geometry simultaneously, namely a rotation-equivariant CNN with decomposed convolutional filters (RotDCF). This decomposition facilitates computing the joint convolution, which is proved to be necessary for the group equivariance. It significantly reduces the model size and computational complexity while preserving performance, and truncation of the bases expansion serves implicitly to regularize the filters. On datasets involving in-plane and out-of-plane object rotations, RotDCF deep features demonstrate greater robustness and interpretability than regular CNNs. The stability of the equivariant representation to input variations is also proved theoretically under generic assumptions on the filters in the decomposed form. The RotDCF framework can be extended to groups other than rotations, providing a general approach which achieves both group equivariance and representation stability at a reduced model size.

References in corpus (11)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
• SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
• Exploiting Linear Structure Within Convolutional Networks for Efficient Evaluation
• Compressing Neural Networks with the Hashing Trick
• Speeding up Convolutional Neural Networks with Low Rank Expansions
• Convolutional neural networks with low-rank regularization
• Speeding-up Convolutional Neural Networks Using Fine-tuned CP-Decomposition
• DCFNet: Deep Neural Network with Decomposed Convolutional Filters
• Oriented Response Networks
• Learning Steerable Filters for Rotation Equivariant CNNs

Cited by in corpus (6)

• On Learning Sets of Symmetric Elements
• Attentive Group Equivariant Convolutional Networks
• Translation Insensitive CNNs
• ChaRRNets: Channel Robust Representation Networks for RF Fingerprinting
• Group Equivariant Stand-Alone Self-Attention For Vision
• FILTRA: Rethinking Steerable CNN by Filter Transform