README.md

SCORE: Scene Context Matters
in Open-Vocabulary Remote Sensing Instance Segmentation

This repository contains code for our ICCV2025 Highlight✨ paper:

SCORE: Scene Context Matters in Open-Vocabulary Remote Sensing Instance Segmentation
Shiqi Huang, Shuting He, Huaiyuan Qin, Bihan Wen
ICCV 2025 (Highlight)

Framework

Installation

Please see Installation Instructions.

Datasets

Download datasets for open-vocabulary remote sensing instance segmentation from Hugging Face .

Getting Started with SCORE

Please see Getting Started with Detectron2 for full usage.

Training

Download RS-CLIP ckpt from RemoteCLIP and put it under SCORE/RemoteCLIP/RemoteCLIP-ViT-L-14.pt.

python train_net.py --num-gpus 1 \
  --config-file configs/score_isaid_instances.yaml

python train_net.py --num-gpus 1 \
  --config-file configs/score_sior_instances.yaml

Inference

bash eval_isaid.sh

bash eval_sior.sh

Acknowledgement

This project is based on FC-CLIP. Many thanks to the authors for their great work!

BibTeX

Please consider to cite SCORE if it helps your research.

@inproceedings{SCORE,
  title={SCORE: Scene Context Matters in Open-Vocabulary Remote Sensing Instance Segmentation},
  author={Huang, Shiqi and He, Shuting and Qin, Huaiyuan and Wen, Bihan},
  booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
  pages={12559--12569},
  year={2025}
}