Research map / Continual learning / Class-incremental
Unlocking Positive Transfer in Incrementally Learning Surgical Instruments: A Self-reflection Hierarchical Prompt Framework
Yu Zhu, Kang Li, Zheng Li, Pheng-Ann Heng
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 prompt framework (HPPT + SRS) to SAM (ablation) | SAM ViT-H · EndoVis17->18 nephrectomy | IoU (All): 48.15 -> 59.44 | Improved |
| add hierarchical prompt parsing tree (HPPT) (ablation) | SAM ViT-H · EndoVis17->18 nephrectomy | IoU (All): 48.15 -> 57.89 | Improved |
| add self-reflection refining strategy (SRS) on top of HPPT (ablation) | SAM + HPPT · EndoVis17->18 nephrectomy | IoU (All): 57.89 -> 59.44 | Improved |
| add SRS on top of HPPT (ablation) | SAM + HPPT · EndoVis17->18 nephrectomy new classes | IoU: 22.16 -> 22.30 | No clear effect |
| add SRS on top of HPPT (ablation) | SAM + HPPT · EndoVis17->18 nephrectomy old classes | IoU: 65.29 -> 68.52 | Improved |
| add HPPT to SAM for old classes (ablation) | SAM ViT-H · EndoVis17->18 nephrectomy old classes | IoU: 21.42 -> 65.29 | Improved |
| add HPPT to SAM for regular classes (ablation) | SAM ViT-H · EndoVis17->18 nephrectomy regular classes | IoU: 69.11 -> 69.23 | No clear effect |
| apply full framework to DeepLabv3+ backbone (ablation) | DeepLabv3+ · EndoVis17->18 nephrectomy | IoU (All): 27.62 -> 40.92 | Improved |
| use full framework instead of best prior CIL method (CAT-SD), same backbone | DeepLabv3+ · EndoVis17->18 nephrectomy | IoU (All): 35.09 -> 40.92 | Improved |
| use framework instead of CAT-SD on cholecystectomy | DeepLabv3+ · CholecSeg8k->M2CAI-Seg | IoU (All): 48.93 -> 53.71 | Improved |
| use framework for positive backward transfer | DeepLabv3+ · EndoVis17->18 nephrectomy | BWT: -0.88 -> 6.22 | Improved |
| add framework on SAM backbone for cholecystectomy FWT | SAM ViT-H · CholecSeg8k->M2CAI-Seg | FWT: 0.69 (Ours SAM) vs 5.70 (Ours DeepLabv3+) | Mixed |
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.
- CoinSeg — CoinSeg: Contrast Inter- and Intra- Class Representations for Incremental Segmentation (ICCV 2023) · better on 1 of 1 shared result
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
- SurgicalSAM: Efficient Class Promptable Surgical Instrument Segmentation (AAAI 2024 · similarity 0.88)
- Procedure-Aware Surgical Video-language Pretraining with Hierarchical Knowledge Augmentation (NeurIPS 2024 · similarity 0.87)
- Where It Moves, It Matters: Referring Surgical Instrument Segmentation via Motion (AAAI 2026 · similarity 0.86)
- Targeted Forgetting of Image Subgroups in CLIP Models (CVPR 2025 · similarity 0.85)
- Knowledge Transfer from Vision Foundation Models for Efficient Training of Small Task-specific Models (ICML 2024 · similarity 0.84)
- SurgCoT: Advancing Spatiotemporal Reasoning in Surgical Videos through a Chain-of-Thought Benchmark (CVPR 2026 · similarity 0.84)
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.
