Multi-Scale Iterative Refinement Network for RGB-D Salient Object Detection

TL;DR

This paper introduces a multi-scale iterative refinement network with an attention-based fusion module for RGB-D salient object detection, effectively addressing cross-modal correlation and multi-scale feature integration.

Contribution

It proposes a novel top-down and bottom-up refinement architecture combined with an attention-based fusion module for improved RGB-D salient object detection.

Findings

Effective multi-scale feature utilization demonstrated

Significant performance improvements on seven datasets

Robust cross-modal fusion achieved

Abstract

The extensive research leveraging RGB-D information has been exploited in salient object detection. However, salient visual cues appear in various scales and resolutions of RGB images due to semantic gaps at different feature levels. Meanwhile, similar salient patterns are available in cross-modal depth images as well as multi-scale versions. Cross-modal fusion and multi-scale refinement are still an open problem in RGB-D salient object detection task. In this paper, we begin by introducing top-down and bottom-up iterative refinement architecture to leverage multi-scale features, and then devise attention based fusion module (ABF) to address on cross-modal correlation. We conduct extensive experiments on seven public datasets. The experimental results show the effectiveness of our devised method