Research map / Detection and segmentation / Open-set and few-shot detection
MultiverSeg: Scalable Interactive Segmentation of Biomedical Imaging Datasets with In-Context Guidance
Hallee E. Wong, Jose Javier Gonzalez Ortiz, John Guttag, Adrian V. Dalca
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 context set of previously segmented images to interactive segmentation | MultiverSeg · Unseen biomedical tasks (clicks) | total clicks: 36.41 ± 1.33% fewer clicks than ScribblePrompt | Improved |
| add context set of previously segmented images to interactive segmentation | MultiverSeg · Unseen biomedical tasks (scribbles) | scribble steps: 25.26 ± 1.80% fewer scribble steps than ScribblePrompt | Improved |
| threshold predictions at 0.5 before adding to context set (ablation) | MultiverSeg and SP+UVS · Unseen biomedical tasks | accuracy of subsequent predictions | Improved |
| use ground truth labels instead of predicted segmentations in context set (ablation) | MultiverSeg · Unseen biomedical tasks | interactions to 90% Dice | Improved |
| bootstrap UniverSeg with its own predictions as context | UniverSeg · Unseen biomedical tasks | Dice: 48.89 ± 1.87 Dice | Hurt |
| add context set at small sizes for click prompts | MultiverSeg · BUID (clicks) | interactions: underperforms ScribblePrompt until context set has >= 5 examples | Hurt |
| add context set when scribbles are the prompt | MultiverSeg · BUID (scribbles) | interactions | No clear effect |
| double context set size versus one more scribble step (ablation) | MultiverSeg · Unseen tasks (ground truth context) | Dice | Mixed |
| train on more datasets with wider CrossBlocks and normalization versus UniverSeg | MultiverSeg · In-context segmentation, all context sizes | Dice: 67 vs. 53 datasets; 256 vs. 64 features | Improved |
Later papers that compared with this method
1 later accepted paper reports MultiverSeg as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- K-Prism: A Knowledge-Guided and Prompt Integrated Universal Medical Image Segmentation Model (ICLR 2026 · 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
- Interactive Medical Image Segmentation: A Benchmark Dataset and Baseline (CVPR 2025 · similarity 0.88)
- UniPro: Unified Multi-Mode Medical Image Segmentation from 2D Images to 3D Volumes via Propagation (NeurIPS 2026 · similarity 0.86)
- Rethinking Interactive Image Segmentation with Low Latency High Quality and Diverse Prompts (CVPR 2024 · similarity 0.85)
- SegVol: Universal and Interactive Volumetric Medical Image Segmentation (NeurIPS 2024 · similarity 0.85)
- CAT: Coordinating Anatomical-Textual Prompts for Multi-Organ and Tumor Segmentation (NeurIPS 2024 · similarity 0.85)
- Scaling up Image Segmentation across Data and Tasks (CVPR 2025 · similarity 0.85)
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.
