Multi-scale Edge-based U-shape Network for Salient Object Detection

TL;DR

This paper introduces MEUN, a multi-scale edge-based U-shape network that enhances salient object detection by improving boundary accuracy and feature integration, outperforming existing methods on benchmark datasets.

Contribution

The paper proposes a novel multi-scale edge-based U-shape network with embedded edge modules and an additional down-sampling module for improved salient object detection.

Findings

Outperforms 15 state-of-the-art methods on benchmark datasets.

Effectively improves boundary sharpness and location accuracy.

Demonstrates robustness and reliability across multiple datasets.

Abstract

Deep-learning based salient object detection methods achieve great improvements. However, there are still problems existing in the predictions, such as blurry boundary and inaccurate location, which is mainly caused by inadequate feature extraction and integration. In this paper, we propose a Multi-scale Edge-based U-shape Network (MEUN) to integrate various features at different scales to achieve better performance. To extract more useful information for boundary prediction, U-shape Edge Network modules are embedded in each decoder units. Besides, the additional down-sampling module alleviates the location inaccuracy. Experimental results on four benchmark datasets demonstrate the validity and reliability of the proposed method. Multi-scale Edge based U-shape Network also shows its superiority when compared with 15 state-of-the-art salient object detection methods.