MERLIN: Multimodal Embedding Refinement via LLM-based Iterative Navigation for Text-Video Retrieval-Rerank Pipeline

TL;DR

MERLIN is a training-free, LLM-based pipeline that iteratively refines text-video retrieval by incorporating user perspective, significantly improving retrieval accuracy on multiple datasets.

Contribution

Introduces MERLIN, a novel LLM-driven, training-free method for iterative query refinement in text-video retrieval, emphasizing user perspective and dynamic feedback.

Findings

Significantly improves Recall@1 on MSR-VTT, MSVD, and ActivityNet datasets.

Outperforms existing retrieval systems in accuracy.

Demonstrates the effectiveness of LLMs in multimodal retrieval refinement.

Abstract

The rapid expansion of multimedia content has made accurately retrieving relevant videos from large collections increasingly challenging. Recent advancements in text-video retrieval have focused on cross-modal interactions, large-scale foundation model training, and probabilistic modeling, yet often neglect the crucial user perspective, leading to discrepancies between user queries and the content retrieved. To address this, we introduce MERLIN (Multimodal Embedding Refinement via LLM-based Iterative Navigation), a novel, training-free pipeline that leverages Large Language Models (LLMs) for iterative feedback learning. MERLIN refines query embeddings from a user perspective, enhancing alignment between queries and video content through a dynamic question answering process. Experimental results on datasets like MSR-VTT, MSVD, and ActivityNet demonstrate that MERLIN substantially improves Recall@1, outperforming existing systems and confirming the benefits of integrating LLMs into multimodal retrieval systems for more responsive and context-aware multimedia retrieval.