mlsanityData to 14 Oct 2026

Benchmarking Composed Image Retrieval for Applied Earth Observation

arXiv 2605.24442 ↗First posted 23 May 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Remote sensing composed image retrieval (RSCIR) enables search in large satellite image archives using composed queries that combine a reference image with a textual modifier. Although RSCIR offers a flexible interface for expressing targeted retrieval intent, the transferability of modern composition methods to Earth observation (EO) imagery and their relevance to operational EO workflows remain underexplored. We address this gap through a unified benchmark and an application-oriented study. First, we systematically adapt and evaluate representative composed image retrieval methods with six vision-language backbones on PatternCom under a standardized protocol, analyzing their behavior across backbones, composition strategies, and query types. Second, we introduce xView2-CIR, a change-centric dataset for disaster and damage monitoring, where retrieval is conditioned on scene identity and a target post-event state. Our results show that training-free composition methods provide strong and scalable baselines for EO retrieval, while change-centric retrieval presents different challenges from attribute-based retrieval, particularly due to the need to preserve scene identity. Overall, this study establishes a practical benchmark for RSCIR and positions composed retrieval as a complementary tool for remote sensing image retrieval, archive exploration, and change analysis. The dataset and code are available at https://github.com/billpsomas/rscir.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 39 of its 100 references that are in the map. First posted on arXiv 23 May 2026.

Builds on

39 of 100 references.

References · 39

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  4. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 13,028
  5. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,639
  6. 2021Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervisioncited by 3,864
  7. 2023Sigmoid Loss for Language Image Pre-Trainingcited by 3,109
  8. 2021Open-vocabulary Object Detection via Vision and Language Knowledge Distillationcited by 906
  9. 2017Fine-tuning CNN Image Retrieval with No Human Annotationcited by 499
  10. 2016Large-Scale Image Retrieval with Attentive Deep Local Featurescited by 238
  11. 2018Composing Text and Image for Image Retrieval - An Empirical Odysseycited by 236
  12. 2023Zero-Shot Composed Image Retrieval with Textual Inversioncited by 176
27 more
  1. 2023Pic2Word: Mapping Pictures to Words for Zero-shot Composed Image Retrievalcited by 156
  2. 2021ZeroCap: Zero-Shot Image-to-Text Generation for Visual-Semantic Arithmeticcited by 150
  3. 2016Deep Image Retrieval: Learning global representations for image searchcited by 147
  4. 2018xView: Objects in Context in Overhead Imagerycited by 146
  5. 2017Automatic Spatially-aware Fashion Concept Discoverycited by 134
  6. 2020Disentangled Non-Local Neural Networkscited by 132
  7. 2023SkyScript: A Large and Semantically Diverse Vision-Language Dataset for Remote Sensingcited by 125
  8. 2024MagicLens: Self-Supervised Image Retrieval with Open-Ended Instructionscited by 119
  9. 2023Vision-by-Language for Training-Free Compositional Image Retrievalcited by 103
  10. 2017PatternNet: A Benchmark Dataset for Performance Evaluation of Remote Sensing Image Retrievalcited by 97
  11. 2021BigEarthNet-MM: A Large Scale Multi-Modal Multi-Label Benchmark Archive for Remote Sensing Image Classification and Retrievalcited by 92
  12. 2022Beyond Cross-view Image Retrieval: Highly Accurate Vehicle Localization Using Satellite Imagecited by 88
  13. 2023CompoDiff: Versatile Composed Image Retrieval With Latent Diffusioncited by 85
  14. 2022ARTEMIS: Attention-based Retrieval with Text-Explicit Matching and Implicit Similaritycited by 83
  15. 2022Exploring a Fine-Grained Multiscale Method for Cross-Modal Remote Sensing Image Retrievalcited by 77
  16. 2019Adversarial Representation Learning for Text-to-Image Matchingcited by 65
  17. 2022Probabilistic Compositional Embeddings for Multimodal Image Retrievalcited by 29
  18. 2025LPOSS: Label Propagation Over Patches and Pixels for Open-vocabulary Semantic Segmentationcited by 21
  19. 2024Composed Image Retrieval for Remote Sensingcited by 14
  20. 2019Enhancing Remote Sensing Image Retrieval with Triplet Deep Metric Learning Networkcited by 12
  21. 2025Instance-Level Composed Image Retrievalcited by 12
  22. 2021Informative and Representative Triplet Selection for Multilabel Remote Sensing Image Retrievalcited by 9
  23. 2021A Novel Graph-Theoretic Deep Representation Learning Method for Multi-Label Remote Sensing Image Retrievalcited by 7
  24. 2016Visual descriptors for content-based retrieval of remote sensing imagescited by 6
  25. 2016Learning Low Dimensional Convolutional Neural Networks for High-Resolution Remote Sensing Image Retrievalcited by 6
  26. 2019Aggregated Deep Local Features for Remote Sensing Image Retrievalcited by 5
  27. 2020Cross-View Image Retrieval -- Ground to Aerial Image Retrieval through Deep Learningcited by 2

Built on by

Not cited yet.