Research map / Detection and segmentation / Open-set and few-shot detection
OpenRSD: Towards Open-prompts for Object Detection in Remote Sensing Images
Ziyue Huang, Yongchao Feng, Ziqi Liu, Shuai Yang, Qingjie Liu, Yunhong Wang
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 fusion head to alignment head (ablation) | OpenRSD · DIOR-R | AP50: 71.5 -> 73.1 | Improved |
| add fusion head to alignment head (ablation) | OpenRSD · HRSC2016 | AP50: 84.3 -> 73.1 | Hurt |
| add class embeddings to fusion head (ablation) | OpenRSD · 5-dataset average | AP50: 61.8 -> 64.3 | Improved |
| add class embeddings to fusion head (ablation) | OpenRSD · DIOR-R | AP50: 73.1 -> 72.9 | Hurt |
| infer with fusion head instead of alignment head (ablation) | OpenRSD · 7-dataset average | AP50: 67.0 -> 68.2 | Improved |
| add pretraining before fine-tuning (ablation) | OpenRSD · 5-dataset average | AP50: 64.0 -> 65.5 | Improved |
| add self-training with full-parameter fine-tuning (ablation) | OpenRSD · 5-dataset average | AP50: 65.5 -> 66.9 | Improved |
| freeze backbone during self-training instead of full-parameter (ablation) | OpenRSD · 5-dataset average | AP50: 66.9 -> 65.7 | Hurt |
| add 250k unlabeled Million-AID images to self-training (ablation) | OpenRSD · 5-dataset average | AP50: 65.5 -> 65.3 | No clear effect |
| increase image prompts from 1 to 10 (ablation) | OpenRSD · DOTA-v2.0 (OBB) | AP50: 65.7 -> 68.2 | Improved |
| increase text prompts from 1 to 20 (ablation) | OpenRSD · DOTA-v2.0 (OBB) | AP50: 70.1 -> 70.1 | No clear effect |
| use OpenRSD instead of YOLO-World-L | YOLO-World-L · DOTA-v2.0 (HBB) | AP50: 58.0 -> 71.8 | Improved |
Later papers that compared with this method
1 later accepted paper reports OpenRSD as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- SLIP-RS: Structured-Attribute Language-Image Pre-Training for Remote Sensing Object Detection (ICML 2026 · better on 4 of 4 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
- Locate Anything on Earth: Advancing Open-Vocabulary Object Detection for Remote Sensing Community (AAAI 2025 · similarity 0.89)
- Simple Image-Level Classification Improves Open-Vocabulary Object Detection (AAAI 2024 · similarity 0.88)
- OW-OVD: Unified Open World and Open Vocabulary Object Detection (CVPR 2025 · similarity 0.87)
- InstructSAM: A Training-free Framework for Instruction-Oriented Remote Sensing Object Recognition (NeurIPS 2025 · similarity 0.87)
- Semi-supervised Open-World Object Detection (AAAI 2024 · similarity 0.87)
- How to Evaluate the Generalization of Detection? A Benchmark for Comprehensive Open-Vocabulary Detection (AAAI 2024 · similarity 0.86)
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.
