Research map / Detection and segmentation / CNN features
RGB-D Salient Object Detection via 3D Convolutional Neural Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use progressive RGB-depth fusion with 3D convolutions in the encoder and decoder | RD3D · NJU2K, NLPR, STERE, RGBD135, DUTLF-D, SIP | S-measure | Helped |
| Add rich back-projection paths to the decoder (ablation) | RD3D · NLPR, NJU2K, STERE, SIP | S-measure: .918 -> .930; .913 -> .916; .906 -> .911; .878 -> .885 | Helped |
| Use channel-modality attention instead of channel-only squeeze-excitation attention (ablation) | RD3D · NLPR, NJU2K, STERE, SIP | S-measure: .921 -> .930; .914 -> .916; .907 -> .911; .866 -> .885 | Helped |
| Move CMA modules from the decoder to the encoder (ablation) | RD3D · NLPR, NJU2K, STERE, SIP | S-measure: .930 -> .931; .916 -> .920; .911 -> .908; .885 -> .883 | Mixed |
| Add CMA and RBPP to the plain 3D UNet baseline (ablation) | RD3D · NLPR, NJU2K, STERE, SIP | S-measure: .913 -> .930; .906 -> .916; .897 -> .911; .873 -> .885 | Helped |
Closest papers by meaning
- Calibrated RGB-D Salient Object Detection (CVPR 2021 · similarity 0.92)
- Deep RGB-D Saliency Detection With Depth-Sensitive Attention and Automatic Multi-Modal Fusion (CVPR 2021 · similarity 0.92)
- Self-Supervised Pretraining for RGB-D Salient Object Detection (AAAI 2022 · similarity 0.92)
- Alignment-Free RGB-T Salient Object Detection: A Large-Scale Dataset and Progressive Correlation Network (AAAI 2025 · similarity 0.90)
- DiMSOD: A Diffusion-Based Framework for Multi-Modal Salient Object Detection (AAAI 2025 · similarity 0.89)
- DFormer: Rethinking RGBD Representation Learning for Semantic Segmentation (ICLR 2024 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.