Co-Attentive Equivariant Neural Networks: Focusing Equivariance On Transformations Co-Occurring In Data
arXiv:1911.07849
Abstract
Equivariance is a nice property to have as it produces much more parameter efficient neural architectures and preserves the structure of the input through the feature mapping. Even though some combinations of transformations might never appear (e.g. an upright face with a horizontal nose), current equivariant architectures consider the set of all possible transformations in a transformation group when learning feature representations. Contrarily, the human visual system is able to attend to the set of relevant transformations occurring in the environment and utilizes this information to assist and improve object recognition. Based on this observation, we modify conventional equivariant feature mappings such that they are able to attend to the set of co-occurring transformations in data and generalize this notion to act on groups consisting of multiple symmetries. We show that our proposed co-attentive equivariant neural networks consistently outperform conventional rotation equivariant and rotation & reflection equivariant neural networks on rotated MNIST and CIFAR-10.
arXiv commentProceedings of the 8th International Conference on Learning Representations (ICLR), 2020
References in corpus (8)
- Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
- Striving for Simplicity: The All Convolutional Net
- Group Equivariant Convolutional Networks
- Rotation equivariant vector field networks
- Long Short-Term Memory-Networks for Machine Reading
- Gauge Equivariant Convolutional Networks and the Icosahedral CNN
- Deep Scale-spaces: Equivariance Over Scale
- Scale equivariance in CNNs with vector fields