Affine Self Convolution

Attention mechanisms, and most prominently self-attention, are a powerful building block for processing not only text but also images. These provide a parameter efficient method for aggregating inputs. We focus on self-attention in vision models, and we combine it with convolution, which as far as w...

Full description

Saved in:
Bibliographic Details
Main Authors Diaconu, Nichita, Worrall, Daniel E
Format Journal Article
LanguageEnglish
Published 18.11.2019
Subjects
Computer Science - Computer Vision and Pattern Recognition
Online AccessGet full text

Cover

More Information
Summary:Attention mechanisms, and most prominently self-attention, are a powerful building block for processing not only text but also images. These provide a parameter efficient method for aggregating inputs. We focus on self-attention in vision models, and we combine it with convolution, which as far as we know, are the first to do. What emerges is a convolution with data dependent filters. We call this an Affine Self Convolution. While this is applied differently at each spatial location, we show that it is translation equivariant. We also modify the Squeeze and Excitation variant of attention, extending both variants of attention to the roto-translation group. We evaluate these new models on CIFAR10 and CIFAR100 and show an improvement in the number of parameters, while reaching comparable or higher accuracy at test time against self-trained baselines.
DOI:10.48550/arxiv.1911.07704