Research map / Detection and segmentation / Open-set and few-shot detection
Test-Time Multi-Prompt Adaptation for Open-Vocabulary Remote Sensing Image Segmentation
Ting Yang, Qilong Wang, Qibin Hou, Qinghua Hu
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 Cat-Prompt generated descriptions with dataset description (ablation) | SegEarth-OV · Vaihingen | mIoU: 29.1 -> 35.9 | Improved |
| add dataset description to Cat-Prompt (ablation) | SegEarth-OV · Vaihingen | mIoU: 32.1 -> 35.9 | Improved |
| add VGTA test-time adaptation on Cat-Prompt (ablation) | SegEarth-OV · Vaihingen | mIoU: 35.9 -> 43.4 | Improved |
| build prompt bias from high-confidence visual features instead of zero-initialized vector (ablation) | VGTA · Vaihingen | mIoU: 41.3 -> 43.4 | Improved |
| increase descriptions per category from 5 to 7 (ablation) | Cat-Prompt · Vaihingen | mIoU: 35.9 -> 33.6 | Hurt |
| use 5 generated descriptions instead of class names only (ablation) | SegEarth-OV · Potsdam | mIoU: 47.1 -> 48.9 | Improved |
| increase descriptions from 1 to 3 (ablation) | Cat-Prompt · Vaihingen | mIoU: 32.5 -> 32.1 | No clear effect |
| apply TMPA to CASS | CASS · WHUAerial | mIoU: 38.9 -> 57.4 | Improved |
| apply TMPA to ClearCLIP | ClearCLIP · Vaihingen | mIoU: 27.3 -> 32.0 | Improved |
| apply TMPA to SegEarth-OV | SegEarth-OV · 8 multi-class datasets (avg) | mIoU: 39.1 -> 43.7 | Improved |
| use 896x896 input resolution | TMPA · WHUAerial | IoU: 55.6 -> 57.8 | Improved |
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.
- SegEarth-OV — SegEarth-OV: Towards Training-Free Open-Vocabulary Segmentation for Remote Sensing Images (CVPR 2025) · better on 14 of 14 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
- Towards Open-Vocabulary Remote Sensing Image Semantic Segmentation (AAAI 2025 · similarity 0.90)
- Test-Time Adaptation of Vision-Language Models for Open-Vocabulary Semantic Segmentation (NeurIPS 2025 · similarity 0.89)
- TEOChat: A Large Vision-Language Assistant for Temporal Earth Observation Data (ICLR 2025 · similarity 0.87)
- Quality-Driven Curation of Remote Sensing Vision-Language Data via Learned Scoring Models (NeurIPS 2025 · similarity 0.87)
- SkySense-O: Towards Open-World Remote Sensing Interpretation with Vision-Centric Visual-Language Modeling (CVPR 2025 · similarity 0.87)
- LISAt: Language-Instructed Segmentation Assistant for Satellite Imagery (NeurIPS 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.
