GOAL: A Challenging Knowledge-grounded Video Captioning Benchmark for Real-time Soccer Commentary Generation

TL;DR

GOAL introduces a new benchmark dataset for knowledge-grounded video captioning in soccer, emphasizing the challenge of generating detailed, knowledge-informed commentary for sports videos.

Contribution

The paper presents GOAL, a large-scale, challenging dataset for knowledge-grounded video captioning, and evaluates existing methods to highlight the task's difficulty and future directions.

Findings

Existing methods struggle with the task complexity.

The dataset enables research on knowledge integration in video captioning.

Baseline results indicate significant room for improvement.

Abstract

Despite the recent emergence of video captioning models, how to generate vivid, fine-grained video descriptions based on the background knowledge (i.e., long and informative commentary about the domain-specific scenes with appropriate reasoning) is still far from being solved, which however has great applications such as automatic sports narrative. In this paper, we present GOAL, a benchmark of over 8.9k soccer video clips, 22k sentences, and 42k knowledge triples for proposing a challenging new task setting as Knowledge-grounded Video Captioning (KGVC). Moreover, we conduct experimental adaption of existing methods to show the difficulty and potential directions for solving this valuable and applicable task. Our data and code are available at https://github.com/THU-KEG/goal.