Region-R1: Reinforcing Query-Side Region Cropping for Multi-Modal Re-Ranking

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hu, Chan-Wei, Tu, Zhengzhong
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914454917087232
author Hu, Chan-Wei
Tu, Zhengzhong
author_facet Hu, Chan-Wei
Tu, Zhengzhong
contents Multi-modal retrieval-augmented generation (MM-RAG) relies heavily on re-rankers to surface the most relevant evidence for image-question queries. However, standard re-rankers typically process the full query image as a global embedding, making them susceptible to visual distractors (e.g., background clutter) that skew similarity scores. We propose Region-R1, a query-side region cropping framework that formulates region selection as a decision-making problem during re-ranking, allowing the system to learn to retain the full image or focus only on a question-relevant region before scoring the retrieved candidates. Region-R1 learns a policy with a novel region-aware group relative policy optimization (r-GRPO) to dynamically crop a discriminative region. Across two challenging benchmarks, E-VQA and InfoSeek, Region-R1 delivers consistent gains, achieving state-of-the-art performances by increasing conditional Recall@1 by up to 20%. These results show the great promise of query-side adaptation as a simple but effective way to strengthen MM-RAG re-ranking.
format Preprint
id arxiv_https___arxiv_org_abs_2604_05268
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Region-R1: Reinforcing Query-Side Region Cropping for Multi-Modal Re-Ranking
Hu, Chan-Wei
Tu, Zhengzhong
Computer Vision and Pattern Recognition
Artificial Intelligence
Computation and Language
Multi-modal retrieval-augmented generation (MM-RAG) relies heavily on re-rankers to surface the most relevant evidence for image-question queries. However, standard re-rankers typically process the full query image as a global embedding, making them susceptible to visual distractors (e.g., background clutter) that skew similarity scores. We propose Region-R1, a query-side region cropping framework that formulates region selection as a decision-making problem during re-ranking, allowing the system to learn to retain the full image or focus only on a question-relevant region before scoring the retrieved candidates. Region-R1 learns a policy with a novel region-aware group relative policy optimization (r-GRPO) to dynamically crop a discriminative region. Across two challenging benchmarks, E-VQA and InfoSeek, Region-R1 delivers consistent gains, achieving state-of-the-art performances by increasing conditional Recall@1 by up to 20%. These results show the great promise of query-side adaptation as a simple but effective way to strengthen MM-RAG re-ranking.
title Region-R1: Reinforcing Query-Side Region Cropping for Multi-Modal Re-Ranking
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2604.05268