UniRank: End-to-End Domain-Specific Reranking of Hybrid Text-Image Candidates

TL;DR

UniRank is a novel domain-specific multimodal reranking framework that directly scores hybrid text-image candidates, improving retrieval performance without modality conversion or extensive domain data.

Contribution

It introduces a unified scoring interface and an end-to-end domain adaptation pipeline for effective multimodal reranking in domain-specific scenarios.

Findings

Outperforms state-of-the-art baselines in scientific literature retrieval and patent search.

Improves Recall@1 by 8.9% and 7.3% respectively.

Demonstrates effective cross-modal relevance scoring without modality conversion.

Abstract

Reranking is a critical component in many information retrieval pipelines. Despite remarkable progress in text-only settings, multimodal reranking remains challenging, particularly when the candidate set contains hybrid text and image items. A key difficulty is the modality gap: a text reranker is intrinsically closer to text candidates than to image candidates, leading to biased and suboptimal cross-modal ranking. Vision-language models (VLMs) mitigate this gap through strong cross-modal alignment and have recently been adopted to build multimodal rerankers. However, most VLM-based rerankers encode all candidates as images, and treating text as images introduces substantial computational overhead. Meanwhile, existing open-source multimodal rerankers are typically trained on general-domain data and often underperform in domain-specific scenarios. To address these limitations, we propose UniRank, a VLM-based reranking framework that natively scores and orders hybrid text-image candidates without any modality conversion. Building on this hybrid scoring interface, UniRank provides an end-to-end domain adaptation pipeline that includes: (1) an instruction-tuning stage that learns calibrated cross-modal relevance scoring by mapping label-token likelihoods to a unified scalar score; and (2) a hard-negative-driven preference alignment stage that constructs in-domain pairwise preferences and performs query-level policy optimization through reinforcement learning from human feedback (RLHF). Extensive experiments on scientific literature retrieval and design patent search demonstrate that UniRank consistently outperforms state-of-the-art baselines, improving Recall@1 by 8.9% and 7.3%, respectively.