MS-DETR: Natural Language Video Localization with Sampling Moment-Moment Interaction

TL;DR

MS-DETR introduces a proposal-based, sampling moment-moment interaction model using DETR for efficient and accurate natural language video localization, achieving superior results on multiple datasets.

Contribution

The paper proposes MS-DETR, a novel sampling-based DETR framework that models moment-moment interactions for improved natural language video localization.

Findings

Outperforms existing methods on three public datasets.

Efficient sampling reduces computational complexity.

Effective cross-modal interaction modeling enhances localization accuracy.

Abstract

Given a query, the task of Natural Language Video Localization (NLVL) is to localize a temporal moment in an untrimmed video that semantically matches the query. In this paper, we adopt a proposal-based solution that generates proposals (i.e., candidate moments) and then select the best matching proposal. On top of modeling the cross-modal interaction between candidate moments and the query, our proposed Moment Sampling DETR (MS-DETR) enables efficient moment-moment relation modeling. The core idea is to sample a subset of moments guided by the learnable templates with an adopted DETR (DEtection TRansformer) framework. To achieve this, we design a multi-scale visual-linguistic encoder, and an anchor-guided moment decoder paired with a set of learnable templates. Experimental results on three public datasets demonstrate the superior performance of MS-DETR.