REViT: Roto-reflection Equivariant Convolutional Vision Transformer

2026-06-23

AI Summary

Existing vision transformers struggle to effectively model rotational and reflectional symmetries in images, limiting their performance on orientation-sensitive tasks. This work proposes a discrete roto-reflection equivariant Vision Transformer that, for the first time, efficiently integrates discrete roto-reflection group equivariance into the ViT architecture. By combining convolutional attention mechanisms with equivariant convolutions, the method explicitly preserves rotational, reflectional, and positional symmetries during feature extraction. This approach substantially simplifies the implementation of equivariance while achieving superior performance over existing roto-reflection equivariant neural networks on image classification benchmarks.

Abstract

In this paper, we propose a discrete roto-reflection group equivariant vision transformer with convolutional attention. Roto-reflection equivariant networks preserve the rotational, flip and positional symmetry in feature maps, making them useful for tasks where orientation of the inputs is relevant to the model outputs. In image classification and object detection, most of the studies on roto-reflection equivariant models have focused on using convolutional neural networks rather than vision transformers. In this paper, we examine the challenges involved in achieving equivariance in vision transformers, and we propose a simpler way to implement a discretized roto-reflection group equivariant vision transformer. The experimental results demonstrate that our approach outperforms the existing approaches for developing discrete roto-reflection group equivariant neural networks for image classification.

Problem

• roto-reflection equivariance
• vision transformer
• convolutional attention
• image classification
• symmetry preservation

Innovation

• roto-reflection equivariance
• vision transformer
• convolutional attention
• group equivariant networks
• discrete symmetry