Search papers, labs, and topics across Lattice.
This paper benchmarks composed image retrieval (CIR) methods for Earth observation (EO) imagery, evaluating their performance on a standardized protocol using the PatternCom dataset and introducing a new change-centric dataset, xView2-CIR, focused on disaster monitoring. The study analyzes the behavior of various vision-language backbones and composition strategies, revealing that training-free composition methods offer strong baselines for EO retrieval. It also highlights the unique challenges of change-centric retrieval compared to attribute-based retrieval, particularly in preserving scene identity.
Surprisingly, simple "training-free" methods for combining vision and language beat more complex approaches for satellite image search, offering a scalable baseline for Earth observation.
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.
