Research map / Detection and segmentation / CNN features
Enhancing Prompt Generation with Adaptive Refinement for Camouflaged Object Detection
Xuehan Chen, Guangyu Ren, Tianhong Dai, Tania Stathaki, Hengyan Liu
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 auxiliary sparse embedding (BLIP noun + image embedding) (ablation) | SAM · COD10K | S-measure: 0.810 -> 0.820 | Improved |
| add auxiliary sparse embedding (ablation) | SAM · CHAMELEON | max F-measure: 0.820 -> 0.816 | Hurt |
| add auxiliary sparse embedding (ablation) | SAM · COD10K | MAE: 0.054 -> 0.039 | Improved |
| add mask prompt from full BLIP caption (ablation) | SAM + EmbAux · CAMO | S-measure: 0.805 -> 0.823 | Improved |
| add mask prompt from full BLIP caption (ablation) | SAM + EmbAux · COD10K | max F-measure: 0.751 -> 0.776 | Improved |
| use extracted first noun instead of full caption for prompt generation (ablation) | SAM + EmbAux + P_init · CHAMELEON | S-measure: 0.866 -> 0.882 | Improved |
| use extracted first noun instead of full caption (ablation) | SAM + EmbAux + P_init · COD10K | max F-measure: 0.776 -> 0.789 | Improved |
| use extracted first noun instead of full caption (ablation) | SAM + EmbAux + P_init · CAMO | MAE: 0.081 -> 0.080 | No clear effect |
| add Adaptive Refinement Module to refine the noun prompt (ablation) | SAM + EmbAux + P_init noun · CHAMELEON | S-measure: 0.882 -> 0.932 | Improved |
| add Adaptive Refinement Module (ablation) | SAM + EmbAux + P_init noun · CAMO | max F-measure: 0.807 -> 0.883 | Improved |
| add Adaptive Refinement Module (ablation) | SAM + EmbAux + P_init noun · COD10K | S-measure: 0.846 -> 0.909 | Improved |
| add Adaptive Refinement Module (ablation) | SAM + EmbAux + P_init noun · COD10K | MAE: 0.036 -> 0.020 | Improved |
Closest papers by meaning
- Relax Image-Specific Prompt Requirement in SAM: A Single Generic Prompt for Segmenting Camouflaged Objects (AAAI 2024 · similarity 0.91)
- Multi-modal Segment Anything Model for Camouflaged Scene Segmentation (ICCV 2025 · similarity 0.90)
- Endow SAM with Keen Eyes: Temporal-spatial Prompt Learning for Video Camouflaged Object Detection (CVPR 2024 · similarity 0.90)
- SAM-REF: Introducing Image-Prompt Synergy during Interaction for Detail Enhancement in the Segment Anything Model (CVPR 2025 · similarity 0.87)
- Shift the Lens: Environment-Aware Unsupervised Camouflaged Object Detection (CVPR 2025 · similarity 0.87)
- CamSAM2: Segment Anything Accurately in Camouflaged Videos (NeurIPS 2025 · 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.
