Research map / Detection and segmentation / Open-set and few-shot detection
PromptMoE: A Segmentation Refinement Framework Leveraging Mixture of Experts for Improved Prompting
Stephen Price, Danielle L. Cote, Elke A. Rundensteiner
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 |
|---|---|---|---|
| use PromptMoE instead of SAMRefiner prompting | SAM ViT-H refiner · Mean over 5 benchmarks | mean delta IoU: +5.40 -> +6.24 | Improved |
| PromptMoE vs SegRefiner-HR on BIG | BIG · BIG | delta IoU: +9.55 vs +8.54 | Hurt |
| add bounding box to a single positive point prompt (ablation) | PromptMoE prompt · Mean over 5 benchmarks | mean delta IoU: -18.82 -> -0.71 | Improved |
| add coarse mask prompt (ablation) | PromptMoE prompt · Mean over 5 benchmarks | mean delta IoU: -0.71 -> +4.71 | Improved |
| add negative (exclusion) point (ablation) | PromptMoE prompt · Mean over 5 benchmarks | mean delta IoU: +4.71 -> +5.48 | Improved |
| use five positive points instead of one (ablation) | PromptMoE prompt · Mean over 5 benchmarks | mean delta IoU: +5.48 -> +5.63 | Improved |
| add Prompt-Placement Explorer (ablation) | PromptMoE · Mean over 5 benchmarks | mean delta IoU: +5.63 -> +6.00 | Improved |
| add Dynamic Expert Selector (ablation) | PromptMoE · Mean over 5 benchmarks | mean delta IoU: +6.00 -> +6.24 | Improved |
| DES selecting 2 experts instead of dense fusion of all experts (ablation) | PromptMoE · Mean over 5 benchmarks | mean delta IoU: +6.01 -> +6.24 | Improved |
| change number of DES-selected experts from 2 to 1, 3 or 4 (ablation) | PromptMoE · Mean over 5 benchmarks | mean delta IoU: +6.24 -> +5.90, +5.92, +5.87 | Hurt |
| DES reduces refinement time by activating only a subset of experts (ablation) | PromptMoE · latency | refinement time: reduces refinement time by about 59.9% | Improved |
| restrict DES to six most efficient experts (PromptMoE-Lite) (ablation) | PromptMoE · Mean over 5 benchmarks | mean delta IoU: +6.24 -> +6.16 | 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.
- SAMRefiner — SAMRefiner: Taming Segment Anything Model for Universal Mask Refinement (ICLR 2025) · better on 7 of 7 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
- Leveraging Hallucinations to Reduce Manual Prompt Dependency in Promptable Segmentation (NeurIPS 2024 · similarity 0.89)
- Rethinking Interactive Image Segmentation with Low Latency High Quality and Diverse Prompts (CVPR 2024 · similarity 0.89)
- Prompt-Driven Referring Image Segmentation with Instance Contrasting (CVPR 2024 · similarity 0.89)
- From Pixels to Concepts: Do Segmentation Models Understand What They Segment? (NeurIPS 2026 · similarity 0.88)
- Prompting Diffusion Models for Zero-Shot Instance Segmentation (NeurIPS 2026 · similarity 0.88)
- AoP-SAM: Automation of Prompts for Efficient Segmentation (AAAI 2025 · similarity 0.88)
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.
