RANKVIDEO: Reasoning Reranking for Text-to-Video Retrieval
This addresses the underexplored area of reasoning-based reranking for video retrieval, offering a more efficient and effective solution for video search systems.
The paper tackled the problem of reasoning-based reranking for text-to-video retrieval by introducing RANKVIDEO, which improved retrieval performance by an average of 31% on nDCG@10 on the MultiVENT 2.0 benchmark.
Reranking is a critical component of modern retrieval systems, which typically pair an efficient first-stage retriever with a more expressive model to refine results. While large reasoning models have driven rapid progress in text-centric reranking, reasoning-based reranking for video retrieval remains underexplored. To address this gap, we introduce RANKVIDEO, a reasoning-based reranker for video retrieval that explicitly reasons over query-video pairs using video content to assess relevance. RANKVIDEO is trained using a two-stage curriculum consisting of perception-grounded supervised fine-tuning followed by reranking training that combines pointwise, pairwise, and teacher confidence distillation objectives, and is supported by a data synthesis pipeline for constructing reasoning-intensive query-video pairs. Experiments on the large-scale MultiVENT 2.0 benchmark demonstrate that RANKVIDEO consistently improves retrieval performance within a two-stage framework, yielding an average improvement of 31% on nDCG@10 and outperforming text-only and vision-language reranking alternatives, while more efficient.