Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Cross-Modal Weighting Network for RGB-D Salient Object Detection

About

Depth maps contain geometric clues for assisting Salient Object Detection (SOD). In this paper, we propose a novel Cross-Modal Weighting (CMW) strategy to encourage comprehensive interactions between RGB and depth channels for RGB-D SOD. Specifically, three RGB-depth interaction modules, named CMW-L, CMW-M and CMW-H, are developed to deal with respectively low-, middle- and high-level cross-modal information fusion. These modules use Depth-to-RGB Weighing (DW) and RGB-to-RGB Weighting (RW) to allow rich cross-modal and cross-scale interactions among feature layers generated by different network blocks. To effectively train the proposed Cross-Modal Weighting Network (CMWNet), we design a composite loss function that summarizes the errors between intermediate predictions and ground truth over different scales. With all these novel components working together, CMWNet effectively fuses information from RGB and depth channels, and meanwhile explores object localization and details across scales. Thorough evaluations demonstrate CMWNet consistently outperforms 15 state-of-the-art RGB-D SOD methods on seven popular benchmarks.

Gongyang Li, Zhi Liu, Linwei Ye, Yang Wang, Haibin Ling• 2020

Related benchmarks

TaskDatasetResultRank
RGB-D Salient Object DetectionSTERE
S-measure (Sα)0.905
198
RGB-D Salient Object DetectionNJU2K (test)
S-measure (Sα)0.903
137
RGB-D Salient Object DetectionSIP
S-measure (Sα)0.868
124
RGB-D Salient Object DetectionLFSD
S-measure (Sα)93.4
122
RGBD Saliency DetectionDES
S-measure0.933
102
RGB-D Salient Object DetectionRGBD135
S-measure (Sα)0.934
92
Salient Object DetectionNLPR (test)
F-beta85.7
76
RGB-D Salient Object DetectionNLPR (test)
S-measure (Sα)87.6
71
Saliency DetectionNJUD (test)
MAE0.046
68
RGB-D Saliency DetectionNLPR
Max F-beta0.903
65
Showing 10 of 20 rows

Other info

Code

Follow for update