Attentive Semantic Alignment with Offset-Aware Correlation Kernels

TL;DR

This paper introduces an attentive semantic alignment method with offset-aware correlation kernels that focus on reliable features and handle local transformations, significantly improving semantic correspondence accuracy.

Contribution

The paper proposes a novel attentive alignment approach with offset-aware kernels to filter distractors and model local transformations, advancing semantic correspondence methods.

Findings

Achieves state-of-the-art performance on semantic correspondence benchmarks.

Effectively filters out distractors and handles local transformations.

Demonstrates significant improvement over previous methods.

Abstract

Semantic correspondence is the problem of establishing correspondences across images depicting different instances of the same object or scene class. One of recent approaches to this problem is to estimate parameters of a global transformation model that densely aligns one image to the other. Since an entire correlation map between all feature pairs across images is typically used to predict such a global transformation, noisy features from different backgrounds, clutter, and occlusion distract the predictor from correct estimation of the alignment. This is a challenging issue, in particular, in the problem of semantic correspondence where a large degree of image variations is often involved. In this paper, we introduce an attentive semantic alignment method that focuses on reliable correlations, filtering out distractors. For effective attention, we also propose an offset-aware correlation kernel that learns to capture translation-invariant local transformations in computing correlation values over spatial locations. Experiments demonstrate the effectiveness of the attentive model and offset-aware kernel, and the proposed model combining both techniques achieves the state-of-the-art performance.