XCiT: Cross-Covariance Image Transformers

TL;DR

XCiT introduces a cross-covariance attention mechanism that reduces complexity from quadratic to linear, enabling efficient high-resolution image processing while maintaining high accuracy across multiple vision tasks.

Contribution

The paper proposes XCA, a novel attention mechanism with linear complexity, and integrates it into XCiT, achieving scalable and accurate vision transformers.

Findings

XCiT achieves state-of-the-art results on ImageNet-1k classification.

XCiT performs well on object detection and segmentation benchmarks.

The proposed method scales efficiently to high-resolution images.

Abstract

Following their success in natural language processing, transformers have recently shown much promise for computer vision. The self-attention operation underlying transformers yields global interactions between all tokens ,i.e. words or image patches, and enables flexible modelling of image data beyond the local interactions of convolutions. This flexibility, however, comes with a quadratic complexity in time and memory, hindering application to long sequences and high-resolution images. We propose a "transposed" version of self-attention that operates across feature channels rather than tokens, where the interactions are based on the cross-covariance matrix between keys and queries. The resulting cross-covariance attention (XCA) has linear complexity in the number of tokens, and allows efficient processing of high-resolution images. Our cross-covariance image transformer (XCiT) is built upon XCA. It combines the accuracy of conventional transformers with the scalability of convolutional architectures. We validate the effectiveness and generality of XCiT by reporting excellent results on multiple vision benchmarks, including image classification and self-supervised feature learning on ImageNet-1k, object detection and instance segmentation on COCO, and semantic segmentation on ADE20k.