Divide and Conquer: Structured Reranking for Expert-Level Ecological Image Retrieval

Asmi Kumar, Edward Vendrow, Sara Beery; Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops, 2025, pp. 5175-5184

Abstract

Fine-grained image retrieval in scientific domains such as ecology demands compositional and expert-level reasoning that general-purpose VLMs often lack. In this paper, we introduce a two-stage structured reranking pipeline that augments queries with web-sourced expert knowledge and decomposes them into verifiable subquestions using large multimodal models. Then, images are scored against these subqueries to produce a final relevance ranking. Our method consistently boosts retrieval accuracy, especially on behavioral and contextual queries, while also greatly reducing manual effort, improving interpretability, and advancing automated visual search for scientific research.

Related Material

[pdf] [supp]
[bibtex]
@InProceedings{Kumar_2025_ICCV, author = {Kumar, Asmi and Vendrow, Edward and Beery, Sara}, title = {Divide and Conquer: Structured Reranking for Expert-Level Ecological Image Retrieval}, booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops}, month = {October}, year = {2025}, pages = {5175-5184} }