QuerYD: A video dataset with high-quality text and audio narrations

TL;DR

QuerYD is a large-scale video dataset featuring high-quality text and audio descriptions, designed to advance retrieval and event localization research in video understanding.

Contribution

The paper introduces QuerYD, a novel dataset with detailed, temporally aligned audio and text annotations for videos, supporting retrieval and event localization tasks.

Findings

QuerYD enables training of effective retrieval models.

Models trained on QuerYD outperform previous datasets.

QuerYD facilitates benchmarking in video understanding.

Abstract

We introduce QuerYD, a new large-scale dataset for retrieval and event localisation in video. A unique feature of our dataset is the availability of two audio tracks for each video: the original audio, and a high-quality spoken description of the visual content. The dataset is based on YouDescribe, a volunteer project that assists visually-impaired people by attaching voiced narrations to existing YouTube videos. This ever-growing collection of videos contains highly detailed, temporally aligned audio and text annotations. The content descriptions are more relevant than dialogue, and more detailed than previous description attempts, which can be observed to contain many superficial or uninformative descriptions. To demonstrate the utility of the QuerYD dataset, we show that it can be used to train and benchmark strong models for retrieval and event localisation. Data, code and models are made publicly available, and we hope that QuerYD inspires further research on video understanding with written and spoken natural language.