Research map / Detection and segmentation / Open-set and few-shot detection
Modeling the Distributional Uncertainty for Salient Object Detection Models
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 a deep ensemble rather than a single base model | Salient object detection model · DUT | AUROC: 0.925 -> 0.928 | Helped |
| Apply discrete long-tail model rebalance to continuous saliency prediction | Salient object detection model · DUTS, ECSSD, DUT | AUROC: 0.952 -> 0.918; 0.959 -> 0.940; 0.925 -> 0.873 | Hurt |
| Apply class-weight recalibration after training | Salient object detection model · DUTS, ECSSD, DUT | AUROC: 0.952 -> 0.942; 0.959 -> 0.954; 0.925 -> 0.910 | Hurt |
| Apply class-balanced loss reweighting | Salient object detection model · DUTS, ECSSD, DUT | AUROC: 0.952 -> 0.948; 0.959 -> 0.956; 0.925 -> 0.919 | Hurt |
| Train an auxiliary true-class-probability confidence decoder | Salient object detection model · DUTS, ECSSD, DUT | AUROC: 0.952 -> 0.955; 0.959 -> 0.961; 0.925 -> 0.933 | Helped |
| Clip high penultimate-layer activations using an in-distribution threshold | Salient object detection model · DUTS, ECSSD, DUT | AUROC: 0.952 -> 0.953; 0.959 -> 0.958; 0.925 -> 0.927 | Mixed |
| Adapt the model at test time with teacher-student consistency learning | Salient object detection model · DUTS | maxFβ: 0.849 -> 0.760 | Hurt |
| Use selected test-time augmentations and aggregate their predictions | Salient object detection model · DUTS | AUROC: 0.952 -> 0.953 | Helped |
Closest papers by meaning
- A Causal Debiasing Framework for Unsupervised Salient Object Detection (AAAI 2022 · similarity 0.89)
- Size-invariance Matters: Rethinking Metrics and Losses for Imbalanced Multi-object Salient Object Detection (ICML 2024 · similarity 0.87)
- RUNA: Object-Level Out-of-Distribution Detection via Regional Uncertainty Alignment of Multimodal Representations (AAAI 2025 · similarity 0.87)
- DiMSOD: A Diffusion-Based Framework for Multi-Modal Salient Object Detection (AAAI 2025 · similarity 0.86)
- Epistemic Uncertainty for Generated Image Detection (NeurIPS 2025 · similarity 0.86)
- COUNTS: Benchmarking Object Detectors and Multimodal Large Language Models under Distribution Shifts (CVPR 2025 · similarity 0.85)
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.