RoarGraph: A Projected Bipartite Graph for Efficient Cross-Modal Approximate Nearest Neighbor Search

TL;DR

This paper introduces RoarGraph, a novel graph index designed for efficient cross-modal approximate nearest neighbor search, especially for out-of-distribution queries, significantly improving search speed and accuracy.

Contribution

The paper presents RoarGraph, a projected bipartite graph that leverages query distribution insights to enhance cross-modal ANNS performance for OOD workloads.

Findings

RoarGraph achieves up to 3.56x faster search speed at 90% recall.

It effectively handles OOD queries by addressing their spatial deviation.

The approach outperforms existing methods on modern datasets.

Abstract

Approximate Nearest Neighbor Search (ANNS) is a fundamental and critical component in many applications, including recommendation systems and large language model-based applications. With the advancement of multimodal neural models, which transform data from different modalities into a shared high-dimensional space as feature vectors, cross-modal ANNS aims to use the data vector from one modality (e.g., texts) as the query to retrieve the most similar items from another (e.g., images or videos). However, there is an inherent distribution gap between embeddings from different modalities, and cross-modal queries become Out-of-Distribution (OOD) to the base data. Consequently, state-of-the-art ANNS approaches suffer poor performance for OOD workloads. In this paper, we quantitatively analyze the properties of the OOD workloads to gain an understanding of their ANNS efficiency. Unlike single-modal workloads, we reveal OOD queries spatially deviate from base data, and the k-nearest neighbors of an OOD query are distant from each other in the embedding space. The property breaks the assumptions of existing ANNS approaches and mismatches their design for efficient search. With insights from the OOD workloads, we propose pRojected bipartite Graph (RoarGraph), an efficient ANNS graph index built under the guidance of query distribution. Extensive experiments show that RoarGraph significantly outperforms state-of-the-art approaches on modern cross-modal datasets, achieving up to 3.56x faster search speed at a 90% recall rate for OOD queries.