Conclusion
We developed a lightweight learning framework for efficient RGB-D SOD. First, a hierarchical cross-modal principal network extracts complementary depth features by learning multiscale refinement features. RGB and depth features are alternately input into the CIGM for gradual refinement. A cross-modal interaction strategy can prevent mutual degradation between RGB and depth features and achieve preliminary background noise interference. Then, the GDAM receives prior cues and imports them into a lightweight network that initializes weights. Finally, the saliency map is predicted by the weighted addition of the initial saliency map of the reversed background. Its performance on eight benchmark datasets demonstrates the effectiveness and superiority of the proposed RLLNet in terms of efficiency and robustness.
This is a preview of subscription content, log in via an institution to check access.
References
Zhou W J, Guo Q L, Lei J S, et al. IRFR-Net: interactive recursive feature-reshaping network for detecting salient objects in RGB-D images. IEEE Trans Neural Netw Learn Syst, 2021. doi: https://doi.org/10.1109/TNNLS.2021.3105484
Zhou T, Fan D P, Cheng M M, et al. RGB-D salient object detection: a survey. Comp Visual Media, 2021, 7: 37–69
Zhou W J, Wu J W, Lei J S, et al. Salient object detection in stereoscopic 3D images using a deep convolutional residual autoencoder. IEEE Trans Multimedia, 2021, 23: 3388–3399
Zhang Z, Lin Z, Xu J, et al. Bilateral attention network for RGB-D salient object detection. IEEE Trans Image Process, 2021, 30: 1949–1961
Borji A, Cheng M M, Jiang H, et al. Salient object detection: a benchmark. IEEE Trans Image Process, 2015, 24: 5706–5722
Zhou W J, Lv Y, Lei J S, et al. Global and local-contrast guides content-aware fusion for RGB-D saliency prediction. IEEE Trans Syst Man Cybern Syst, 2021, 51: 3641–3649
Howard A, Sandler M, Chu G, et al. Searching for MobileNetV3. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (CVPR), 2019. 1314–1324
Zhang Y F, Zheng J B, Jia W J, et al. Deep RGB-D saliency detection without depth. IEEE Trans Multimedia, 2022, 24: 755–767
Fan D P, Lin Z, Zhang Z, et al. Rethinking RGB-D salient object detection: models, data sets, and large-scale benchmarks. IEEE Trans Neural Netw Learn Syst, 2021, 32: 2075–2089
Acknowledgements
This work was supported by National Natural Science Foundation of China (Grant Nos. 61502429, 61972357).
Author information
Authors and Affiliations
Corresponding author
Additional information
Supporting information
Appendixes A–C. The supporting information is available online at info.scichina.com and link.springer.com. The supporting materials are published as submitted, without typesetting or editing. The responsibility for scientific accuracy and content remains entirely with the authors.
Electronic supplementary material
Rights and permissions
About this article
Cite this article
Zhou, W., Liu, C., Lei, J. et al. RLLNet: a lightweight remaking learning network for saliency redetection on RGB-D images. Sci. China Inf. Sci. 65, 160107 (2022). https://doi.org/10.1007/s11432-020-3337-9
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s11432-020-3337-9