AntEmbed: Dual-Mode Multimodal Embedding Model for Efficient and Fine-Grained Retrieval
The paper proposes AntEmbed, a dual-mode multimodal embedding model for efficient retrieval and reasoning-enhanced fine-grained matching.
The paper introduces a novel dual-mode approach that combines discriminative and generative embeddings to improve fine-grained matching performance.
Before reading this…
Applications
- →Multimodal retrieval
- →Multimodal matching
To understand this paper, make sure you know these concepts first:
- Multimodal embeddingsfind papers →
- Contrastive learningfind papers →
Abstract
More Like ThisDiscriminative embeddings encode original multimodal inputs efficiently but struggle with fine-grained semantics. Generative embeddings use reasoning to capture query intent and relevance evidence, but incur higher computational costs. We propose AntEmbed, a dual-mode multimodal embedding model that supports efficient retrieval and reasoning-enhanced fine-grained matching in a unified embedding space. First, we construct retrieval-oriented reasoning content that captures query intent and confusion-aware semantic evidence. We treat original and reasoning-enhanced candidate inputs as semantically equivalent views, and use multi-positive contrastive learning with a single embedding token to align discriminative and generative embeddings. Second, cross-mode self-distillation transfers fine-grained semantic information from reasoning-enhanced to original representations through candidate similarity-distribution distillation and query representation alignment. This strengthens discriminative embeddings without requiring additional online reasoning. Both retrieval modes share a single candidate index. On MMEB-v2, AntEmbed-8B achieves an overall score of 81.5 with reasoning-enhanced embeddings, outperforming the strongest baseline by 3.1 points.