# Data-Level Recombination and Lightweight Fusion Scheme for RGB-D Salient Object Detection

arXiv: https://arxiv.org/abs/2009.05102 · HTML: https://mlsanity.com/paper/2009.05102/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 7 Aug 2020.

## Built on by (most cited first, 13 of 13)

- [SwinNet: Swin Transformer drives edge-aware RGB-D and RGB-T salient object detection](https://mlsanity.com/paper/2204.05585/) (2022)
- [Source-free Depth for Object Pop-out](https://mlsanity.com/paper/2212.05370/) (2022)
- [TriTransNet: RGB-D Salient Object Detection with a Triplet Transformer Embedding Network](https://mlsanity.com/paper/2108.03990/) (2021)
- [Edge-aware Guidance Fusion Network for RGB Thermal Scene Parsing](https://mlsanity.com/paper/2112.05144/) (2021)
- [HiDAnet: RGB-D Salient Object Detection via Hierarchical Depth Awareness](https://mlsanity.com/paper/2301.07405/) (2023)
- [GroupTransNet: Group Transformer Network for RGB-D Salient Object Detection](https://mlsanity.com/paper/2203.10785/) (2022)
- [Salient Object Detection via Dynamic Scale Routing](https://mlsanity.com/paper/2210.13821/) (2022)
- [A Novel Long-term Iterative Mining Scheme for Video Salient Object Detection](https://mlsanity.com/paper/2206.09564/) (2022)
- [Dual Swin-Transformer based Mutual Interactive Network for RGB-D Salient Object Detection](https://mlsanity.com/paper/2206.03105/) (2022)
- [Salient Object Detection From Arbitrary Modalities](https://mlsanity.com/paper/2405.03352/) (2024)
- [Depth-guided Texture Diffusion for Image Semantic Segmentation](https://mlsanity.com/paper/2408.09097/) (2024)
- [HODINet: High-Order Discrepant Interaction Network for RGB-D Salient Object Detection](https://mlsanity.com/paper/2307.00954/) (2023)
- [Dual Mutual Learning Network with Global-local Awareness for RGB-D Salient Object Detection](https://mlsanity.com/paper/2501.01648/) (2025)

## Abstract (arXiv)

Existing RGB-D salient object detection methods treat depth information as an independent component to complement its RGB part, and widely follow the bi-stream parallel network architecture. To selectively fuse the CNNs features extracted from both RGB and depth as a final result, the state-of-the-art (SOTA) bi-stream networks usually consist of two independent subbranches; i.e., one subbranch is used for RGB saliency and the other aims for depth saliency. However, its depth saliency is persistently inferior to the RGB saliency because the RGB component is intrinsically more informative than the depth component. The bi-stream architecture easily biases its subsequent fusion procedure to the RGB subbranch, leading to a performance bottleneck. In this paper, we propose a novel data-level recombination strategy to fuse RGB with D (depth) before deep feature extraction, where we cyclically convert the original 4-dimensional RGB-D into DGB, RDB and RGD. Then, a newly lightweight designed triple-stream network is applied over these novel formulated data to achieve an optimal channel-wise complementary fusion status between the RGB and D, achieving a new SOTA performance.