Group Equivariant Stand-Alone Self-Attention For Vision
David W. Romero · Jean-Baptiste Cordonnier
Keywords:
self-attention
transformers
group equivariant transformers
group equivariant self-attention
group equivariance
2021 Poster
Abstract
We provide a general self-attention formulation to impose group equivariance to arbitrary symmetry groups. This is achieved by defining positional encodings that are invariant to the action of the group considered. Since the group acts on the positional encoding directly, group equivariant self-attention networks (GSA-Nets) are steerable by nature. Our experiments on vision benchmarks demonstrate consistent improvements of GSA-Nets over non-equivariant self-attention networks.
Video
Chat is not available.
Successful Page Load