Retrieve Fast, Rerank Smart: Cooperative and Joint Approaches for Improved Cross-Modal Retrieval

TL;DR

This paper introduces a cooperative retrieve-and-rerank framework that enhances cross-modal retrieval by combining efficient bi-encoders for initial retrieval with a cross-encoder for refined ranking, achieving better accuracy and efficiency.

Contribution

It presents a novel fine-tuning approach that transforms pretrained multi-modal models into efficient retrieval systems using shared-weight bi-encoders and cross-encoders.

Findings

Improved retrieval accuracy across multiple benchmarks.

Significant reduction in retrieval latency.

Effective joint fine-tuning of components enhances performance.

Abstract

Current state-of-the-art approaches to cross-modal retrieval process text and visual input jointly, relying on Transformer-based architectures with cross-attention mechanisms that attend over all words and objects in an image. While offering unmatched retrieval performance, such models: 1) are typically pretrained from scratch and thus less scalable, 2) suffer from huge retrieval latency and inefficiency issues, which makes them impractical in realistic applications. To address these crucial gaps towards both improved and efficient cross-modal retrieval, we propose a novel fine-tuning framework that turns any pretrained text-image multi-modal model into an efficient retrieval model. The framework is based on a cooperative retrieve-and-rerank approach which combines: 1) twin networks (i.e., a bi-encoder) to separately encode all items of a corpus, enabling efficient initial retrieval, and 2) a cross-encoder component for a more nuanced (i.e., smarter) ranking of the retrieved small set of items. We also propose to jointly fine-tune the two components with shared weights, yielding a more parameter-efficient model. Our experiments on a series of standard cross-modal retrieval benchmarks in monolingual, multilingual, and zero-shot setups, demonstrate improved accuracy and huge efficiency benefits over the state-of-the-art cross-encoders.