Research map / Detection and segmentation / Open-set and few-shot detection
OFL-SAM2: Prompt SAM2 with Online Few-shot Learner for Efficient Medical Image Segmentation
Meng Lan, Lefei Zhang, Xiaomeng Li
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 online few-shot learner (mapping network) (ablation) | OFL-SAM2 · Synapse-CT (3 volumes) | mean Dice (%): 74.74 -> 78.96 | Improved |
| add adaptive fusion module on top of few-shot learner (ablation) | OFL-SAM2 · Synapse-CT (3 volumes) | mean Dice (%): 78.96 -> 82.52 | Improved |
| add online few-shot learner (mapping network) (ablation) | OFL-SAM2 · PROMISE12 (3 volumes) | mean Dice (%): 83.34 -> 86.21 | Improved |
| add adaptive fusion module on top of few-shot learner (ablation) | OFL-SAM2 · PROMISE12 (3 volumes) | mean Dice (%): 86.21 -> 88.74 | Improved |
| add quality-aware update strategy (gamma 0.8) (ablation) | OFL-SAM2 · Synapse-CT (3 volumes) | mean Dice (%): 81.48 -> 82.52 | Improved |
| add quality-aware update strategy (gamma 0.8) (ablation) | OFL-SAM2 · PROMISE12 (3 volumes) | mean Dice (%): 88.17 -> 88.74 | Improved |
| raise confidence threshold gamma from 0.8 to 0.9 (ablation) | OFL-SAM2 update strategy · Synapse-CT (3 volumes) | mean Dice (%): 82.52 -> 82.04 | Hurt |
| use K=2 similar training images in memory bank instead of 1 (ablation) | OFL-SAM2 · Synapse-CT (3 volumes) | mean Dice (%): 81.86 -> 82.52 | Improved |
| increase similar training images from K=2 to K=3 (ablation) | OFL-SAM2 · Synapse-CT (3 volumes) | mean Dice (%): 82.52 -> 82.21 | No clear effect |
| increase similar training images from K=2 to K=3 (ablation) | OFL-SAM2 · PROMISE12 (3 volumes) | mean Dice (%): 88.74 -> 88.68 | No clear effect |
| reduce training set from 3 to 2 volumes | OFL-SAM2 · Synapse-CT | mean Dice (%): 82.52 -> 81.81 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SurgicalSAM — SurgicalSAM: Efficient Class Promptable Surgical Instrument Segmentation (AAAI 2024) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Enhancing SAM with Efficient Prompting and Preference Optimization for Semi-supervised Medical Image Segmentation (CVPR 2025 · similarity 0.90)
- Unleashing the Potential of SAM for Medical Adaptation via Hierarchical Decoding (CVPR 2024 · similarity 0.89)
- SAM 2: Segment Anything in Images and Videos (ICLR 2025 · similarity 0.88)
- BLO-SAM: Bi-level Optimization Based Finetuning of the Segment Anything Model for Overfitting-Preventing Semantic Segmentation (ICML 2024 · similarity 0.88)
- One-Prompt to Segment All Medical Images (CVPR 2024 · similarity 0.87)
- Towards a Comprehensive, Efficient and Promptable Anatomic Structure Segmentation Model Using 3D Whole-Body CT Scans (AAAI 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.
