Prompting Large Language Models to Reformulate Queries for Moment Localization

TL;DR

This paper explores using large language models to reformulate natural language queries into clearer instructions, improving the accuracy of moment localization in videos by making queries more understandable for models.

Contribution

It introduces a novel approach of leveraging large language models to reformulate complex queries into more effective instructions for video moment localization.

Findings

Reformulated queries improve localization accuracy.

Large language models effectively generate clearer instructions.

Enhanced query understanding aids in complex video scenarios.

Abstract

The task of moment localization is to localize a temporal moment in an untrimmed video for a given natural language query. Since untrimmed video contains highly redundant contents, the quality of the query is crucial for accurately localizing moments, i.e., the query should provide precise information about the target moment so that the localization model can understand what to look for in the videos. However, the natural language queries in current datasets may not be easy to understand for existing models. For example, the Ego4D dataset uses question sentences as the query to describe relatively complex moments. While being natural and straightforward for humans, understanding such question sentences are challenging for mainstream moment localization models like 2D-TAN. Inspired by the recent success of large language models, especially their ability of understanding and generating complex natural language contents, in this extended abstract, we make early attempts at reformulating the moment queries into a set of instructions using large language models and making them more friendly to the localization models.