A 2D Semantic-Aware Position Encoding for Vision Transformers

TL;DR

This paper introduces SaPE², a semantic-aware 2D position encoding for vision transformers that improves their ability to understand spatial relationships by considering content similarity, leading to better generalization and performance.

Contribution

We propose SaPE², a novel position encoding that dynamically adapts based on local content, addressing limitations of traditional fixed position encodings in vision transformers.

Findings

Enhanced model generalization across resolutions and scales

Improved translation equivariance in vision transformers

Better aggregation of features for similar but distant patches

Abstract

Vision transformers have demonstrated significant advantages in computer vision tasks due to their ability to capture long-range dependencies and contextual relationships through self-attention. However, existing position encoding techniques, which are largely borrowed from natural language processing, fail to effectively capture semantic-aware positional relationships between image patches. Traditional approaches like absolute position encoding and relative position encoding primarily focus on 1D linear position relationship, often neglecting the semantic similarity between distant yet contextually related patches. These limitations hinder model generalization, translation equivariance, and the ability to effectively handle repetitive or structured patterns in images. In this paper, we propose 2-Dimensional Semantic-Aware Position Encoding ($\text{SaPE}^2$), a novel position encoding method with semantic awareness that dynamically adapts position representations by leveraging local content instead of fixed linear position relationship or spatial coordinates. Our method enhances the model's ability to generalize across varying image resolutions and scales, improves translation equivariance, and better aggregates features for visually similar but spatially distant patches. By integrating $\text{SaPE}^2$ into vision transformers, we bridge the gap between position encoding and perceptual similarity, thereby improving performance on computer vision tasks.