Research map / Detection and segmentation / Open-set and few-shot detection
Zoom in and Out: A Mixed-Scale Triplet Network for Camouflaged Object Detection
Check 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 HMU (g=6) to the baseline (ablation) | ZoomNet baseline · COD10K-Test | S-measure: 0.812 -> 0.820 | Helped |
| Add SIU to the baseline (ablation) | ZoomNet baseline · COD10K-Test | S-measure: 0.812 -> 0.837 | Helped |
| Add both HMU and SIU (ablation) | ZoomNet baseline · COD10K-Test | S-measure: 0.812 -> 0.843 | Helped |
| Replace proposed modules with standard convolution blocks at similar parameter/FLOP scale (ablation) | ZoomNet baseline · COD10K-Test | S-measure: 0.843 -> 0.815 | Hurt |
| Use mixed 0.5×, 1.0×, and 1.5× inputs with SIU instead of single-scale 1.0× (ablation) | Baseline · CAMO and COD10K average | S-measure: 0.797 -> 0.827 | Helped |
| Set HMU group count to 6 rather than 2, 4, or 8 (ablation) | ZoomNet · COD10K-Test | S-measure: 0.835; 0.836; 0.838; 0.836 | Mixed |
| Add power-form UAL with α=2 (ablation) | ZoomNet without UAL · COD10K-Test | Fβ: 0.743 -> 0.766 | Helped |
| Replace power-form UAL with exponential-form UAL (α=1/4) (ablation) | ZoomNet · COD10K-Test | S-measure: 0.838 -> 0.845 | Helped |
| Replace additive UAL with BCE ambiguity weighting (ablation) | ZoomNet · COD10K-Test | S-measure: 0.838 -> 0.844 | Helped |
| Use cosine scheduling rather than constant λ (ablation) | ZoomNet with UAL · COD10K-Test | S-measure: 0.830 -> 0.838 | Helped |
| Use cosine scheduling rather than linear λ increase from 0 to T (ablation) | ZoomNet with UAL · COD10K-Test | Fβ: 0.760 -> 0.766 | Helped |
| Use zoom-based mixed-scale feature integration and uncertainty-aware training | ZoomNet · CAMO, CHAMELEON, COD10K, and NC4K | S-measure | Helped |
Closest papers by meaning
- Camouflaged Object Detection (CVPR 2020 · similarity 0.92)
- Shift the Lens: Environment-Aware Unsupervised Camouflaged Object Detection (CVPR 2025 · similarity 0.89)
- SdalsNet: Self-Distilled Attention Localization and Shift Network for Unsupervised Camouflaged Object Detection (AAAI 2025 · similarity 0.88)
- CamoDiffusion: Camouflaged Object Detection via Conditional Diffusion Models (AAAI 2024 · similarity 0.87)
- UCOD-DPL: Unsupervised Camouflaged Object Detection via Dynamic Pseudo-label Learning (CVPR 2025 · similarity 0.86)
- Alignment-Free RGB-T Salient Object Detection: A Large-Scale Dataset and Progressive Correlation Network (AAAI 2025 · similarity 0.86)
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.