Bifurcated backbone strategy for RGB-D salient object detection

TL;DR

This paper introduces BBS-Net, a novel RGB-D salient object detection model that uses a bifurcated backbone and depth-enhanced modules to improve multi-modal feature fusion, achieving state-of-the-art results.

Contribution

The paper proposes a bifurcated backbone strategy and depth-enhanced modules for improved multi-modal feature fusion in RGB-D salient object detection.

Findings

BBS-Net outperforms 18 SOTA models on 8 datasets.

Achieves approximately 4% improvement in S-measure over top models.

Demonstrates strong generalization across datasets.

Abstract

Multi-level feature fusion is a fundamental topic in computer vision. It has been exploited to detect, segment and classify objects at various scales. When multi-level features meet multi-modal cues, the optimal feature aggregation and multi-modal learning strategy become a hot potato. In this paper, we leverage the inherent multi-modal and multi-level nature of RGB-D salient object detection to devise a novel cascaded refinement network. In particular, first, we propose to regroup the multi-level features into teacher and student features using a bifurcated backbone strategy (BBS). Second, we introduce a depth-enhanced module (DEM) to excavate informative depth cues from the channel and spatial views. Then, RGB and depth modalities are fused in a complementary way. Our architecture, named Bifurcated Backbone Strategy Network (BBS-Net), is simple, efficient, and backbone-independent. Extensive experiments show that BBS-Net significantly outperforms eighteen SOTA models on eight challenging datasets under five evaluation measures, demonstrating the superiority of our approach ($\sim 4 \%$ improvement in S-measure $vs.$ the top-ranked model: DMRA-iccv2019). In addition, we provide a comprehensive analysis on the generalization ability of different RGB-D datasets and provide a powerful training set for future research.