Research map / Detection and segmentation / CNN features
Self-Supervised Pretraining for RGB-D Salient Object Detection
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 |
|---|---|---|---|
| Add cross-modal joint-consistency features (ablation) | ImageNet-pretrained VGG-16 FPN baseline · Ave-Metric | Fβw: 0.829 -> 0.846 | Helped |
| Add cross-modal joint-difference features (ablation) | CDA Model 2 · Ave-Metric | Fβw: 0.846 -> 0.854 | Helped |
| Add cross-level joint-consistency features (ablation) | ImageNet-pretrained VGG-16 FPN baseline · Ave-Metric | Fβw: 0.829 -> 0.854 | Helped |
| Add cross-level joint-difference features (ablation) | CDA Model 4 · Ave-Metric | Fβw: 0.854 -> 0.861 | Helped |
| Use CDA for both cross-modal and cross-level fusion (ablation) | VGG-16 RGB-D SOD · Ave-Metric | Fβw: 0.829 -> 0.877 | Helped |
| Replace CDA with addition and convolution operations (ablation) | VGG-16 RGB-D SOD · Ave-Metric | Fβw: 0.877 -> 0.829 | Hurt |
| Initialize the encoder with cross-modal auto-encoder pretraining (P1) (ablation) | Randomly initialized VGG-16 RGB-D SOD · Ave-Metric | Fβw: 0.739 -> 0.779 | Helped |
| Add depth-contour estimation pretraining (P2) (ablation) | VGG-16 RGB-D SOD initialized with P1 · Ave-Metric | Fβw: 0.779 -> 0.800 | Helped |
| Add cross-modal CDA (ablation) | VGG-16 RGB-D SOD with P1 · Ave-Metric | Fβw: 0.779 -> 0.812 | Helped |
| Add cross-level CDA (ablation) | VGG-16 RGB-D SOD with P1 · Ave-Metric | Fβw: 0.779 -> 0.820 | Helped |
| Combine P1, P2, cross-modal CDA, and cross-level CDA (ablation) | VGG-16 RGB-D SOD · Ave-Metric | Fβw: 0.792 -> 0.855 | Helped |
| Initialize HAINet with SSL-P1 weights instead of random weights (ablation) | HAINet · Ave-Metric | Fβw: 0.793 -> 0.839 | Helped |
Closest papers by meaning
- RGB-D Salient Object Detection via 3D Convolutional Neural Networks (AAAI 2021 · similarity 0.92)
- Promoting Saliency From Depth: Deep Unsupervised RGB-D Saliency Detection (ICLR 2022 · similarity 0.91)
- DiMSOD: A Diffusion-Based Framework for Multi-Modal Salient Object Detection (AAAI 2025 · similarity 0.89)
- SMR-Net: Semantic-Guided Mutually Reinforcing Network for Cross-Modal Image Fusion and Salient Object Detection (AAAI 2025 · similarity 0.88)
- Alignment-Free RGB-T Salient Object Detection: A Large-Scale Dataset and Progressive Correlation Network (AAAI 2025 · similarity 0.88)
- DFormer: Rethinking RGBD Representation Learning for Semantic Segmentation (ICLR 2024 · similarity 0.87)
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.