Research map / Detection and segmentation / Open-set and few-shot detection
iFS-RCNN: An Incremental Few-Shot Instance Segmenter
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 |
|---|---|---|---|
| Replace sigmoid point estimates with Bayesian probit classification (ablation) | Mask+Sigmoid · COCO FSOD, 1-shot | AP: 3.60 -> 5.18 | Helped |
| Replace sigmoid point estimates with Bayesian probit classification (ablation) | Mask+Sigmoid · COCO FSIS, 1-shot | AP: 3.92 -> 5.15 | Helped |
| Use Monte Carlo sampling instead of the probit approximation (ablation) | Bayesian classifier · COCO FSOD, 1-shot | AP: 5.18 -> 4.52 | Hurt |
| Use Monte Carlo sampling instead of the probit approximation (ablation) | Bayesian classifier · COCO FSIS, 1-shot | AP: 5.15 -> 4.54 | Hurt |
| Add uncertainty-guided box prediction and refinement (ablation) | Mask+Sigmoid · COCO FSOD, 1-shot | AP: 3.60 -> 4.48 | Helped |
| Add uncertainty-guided box prediction and refinement (ablation) | Mask+Sigmoid · COCO FSIS, 1-shot | AP: 3.92 -> 4.84 | Helped |
| Use Gaussian-assumption box uncertainty (ablation) | Mask+Sigmoid · COCO FSOD, 1-shot | AP: 3.60 -> 3.74 | No clear effect |
| Refine boxes without explicitly predicting uncertainty (ablation) | Mask+Sigmoid · COCO FSOD, 1-shot | AP: 3.60 -> 3.87 | No clear effect |
| Add Bayesian probit classification and uncertainty-guided box refinement | Mask+Sigmoid · COCO iFSOD, 10-shot, new classes | AP: 8.04 -> 12.55 | Helped |
| Add Bayesian probit classification and uncertainty-guided box refinement | Mask+Sigmoid · COCO iFSIS, 10-shot, base classes | AP: 35.69 -> 36.32 | Helped |
| Use iFS-RCNN instead of iMTFA | iFSIS · COCO iFSIS, 10-shot, new classes | AP: 5.88 -> 10.06 | Helped |
| Use iFS-RCNN instead of iMTFA | iFSIS · COCO iFSIS, 10-shot, base classes | AP: 21.87 -> 36.32 | Helped |
Closest papers by meaning
- Incremental Few-Shot Instance Segmentation (CVPR 2021 · similarity 0.93)
- FAPIS: A Few-Shot Anchor-Free Part-Based Instance Segmenter (CVPR 2021 · similarity 0.90)
- Incremental Few-Shot Object Detection (CVPR 2020 · similarity 0.89)
- The Devil is in the Object Boundary: Towards Annotation-free Instance Segmentation using Foundation Models (ICLR 2024 · similarity 0.88)
- MaskDiff: Modeling Mask Distribution with Diffusion Probabilistic Model for Few-Shot Instance Segmentation (AAAI 2024 · similarity 0.87)
- Training-Free Dataset Pruning for Instance Segmentation (ICLR 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.