Universal Vision-Language Dense Retrieval: Learning A Unified Representation Space for Multi-Modal Retrieval

TL;DR

This paper introduces UniVL-DR, a unified model for multi-modal retrieval that encodes queries and resources into a shared embedding space, achieving state-of-the-art results on multi-modal benchmarks.

Contribution

It proposes a novel universal embedding optimization and image verbalization techniques to unify multi-modal retrieval in a single model.

Findings

Achieves state-of-the-art on WebQA benchmark.

Outperforms all models on text-text and text-image retrieval.

Demonstrates feasibility of universal multi-modal search.

Abstract

This paper presents Universal Vision-Language Dense Retrieval (UniVL-DR), which builds a unified model for multi-modal retrieval. UniVL-DR encodes queries and multi-modality resources in an embedding space for searching candidates from different modalities. To learn a unified embedding space for multi-modal retrieval, UniVL-DR proposes two techniques: 1) Universal embedding optimization strategy, which contrastively optimizes the embedding space using the modality-balanced hard negatives; 2) Image verbalization method, which bridges the modality gap between images and texts in the raw data space. UniVL-DR achieves the state-of-the-art on the multi-modal open-domain question answering benchmark, WebQA, and outperforms all retrieval models on the two subtasks, text-text retrieval and text-image retrieval. It demonstrates that universal multi-modal search is feasible to replace the divide-and-conquer pipeline with a united model and also benefits single/cross modality tasks. All source codes of this work are available at https://github.com/OpenMatch/UniVL-DR.