Research map / Detection and segmentation / CNN features
Hierarchical Dynamic Filtering Network for RGB-D Salient Object Detection
Youwei Pang, Lihe Zhang, Xiaoqi Zhao, Huchuan Lu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add dynamic dilated pyramid module (DDPM) to single- and dual-modality transport variants (ablation) | HDFNet (VGG-16) · RGB-D SOD AveMetric (8 datasets) | weighted F-measure and MAE: 1.47%, 3.11% and 2.11% relative gain in F_beta^w; 5.01%, 10.29% and 6.77% in MAE | Improved |
| replace the DCM dynamic filter with DDPM (ablation) | HDFNet (VGG-16) · RGB-D SOD AveMetric (8 datasets) | F_ada, weighted F, MAE: 3.91%, 5.60% and 18.07% | Improved |
| add DCM instead of nothing after the transport layer (ablation) | HDFNet Model 6 · RGB-D SOD AveMetric (8 datasets) | weighted F-measure: 0.803 -> 0.777 | Hurt |
| add hybrid enhanced loss (HEL) to Model 7 (ablation) | HDFNet (VGG-16) · RGB-D SOD AveMetric (8 datasets) | weighted F-measure: 0.820 -> 0.857 | Improved |
| add edge loss L_e (ablation) | HDFNet Model 7 · RGB-D SOD AveMetric (8 datasets) | weighted F-measure: 0.820 -> 0.849 | Improved |
| add foreground loss L_f (ablation) | HDFNet Model 7 · RGB-D SOD AveMetric (8 datasets) | weighted F-measure: 0.820 -> 0.827 | Improved |
| add background loss L_b (ablation) | HDFNet Model 7 · RGB-D SOD AveMetric (8 datasets) | weighted F-measure: 0.820 -> 0.836 | Improved |
| add RGB dense transport layer to depth-only transport (ablation) | HDFNet baseline · RGB-D SOD AveMetric (8 datasets) | weighted F-measure: 0.768 -> 0.803 | Improved |
| add depth dense transport layer to baseline (ablation) | HDFNet baseline · RGB-D SOD AveMetric (8 datasets) | weighted F-measure: 0.768 -> 0.768 | No clear effect |
| add HEL to R3Net (ablation) | R3Net · RGB SOD AveMetric (5 datasets) | weighted F-measure: 0.716 -> 0.740 | Improved |
| add HEL to CPD (ablation) | CPD · RGB SOD AveMetric (5 datasets) | weighted F-measure: 0.769 -> 0.792 | Improved |
| add HEL to PoolNet (ablation) | PoolNet · RGB SOD AveMetric (5 datasets) | weighted F-measure: 0.728 -> 0.799 | Improved |
Closest papers by meaning
- RGB-D Salient Object Detection with Cross-Modality Modulation and Selection (ECCV 2020 · similarity 0.92)
- CNN-Based RGB-D Salient Object Detection: Learn, Select, and Fuse (IJCV 2021 · similarity 0.92)
- Specificity-Preserving RGB-D Saliency Detection (ICCV 2021 · similarity 0.91)
- Deep RGB-D Saliency Detection With Depth-Sensitive Attention and Automatic Multi-Modal Fusion (CVPR 2021 · similarity 0.91)
- RGB-D Salient Object Detection via 3D Convolutional Neural Networks (AAAI 2021 · similarity 0.91)
- DiMSOD: A Diffusion-Based Framework for Multi-Modal Salient Object Detection (AAAI 2025 · similarity 0.90)
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.
