Deep soccer captioning with transformer: dataset, semantics-related losses, and multi-level evaluation

TL;DR

This paper introduces a new dataset, a multi-component transformer-based model, and a comprehensive three-level evaluation for generating descriptive captions for soccer videos, significantly improving caption diversity and accuracy.

Contribution

It presents a novel dataset, a multi-modal transformer model, and a triple-level evaluation framework for soccer video captioning, advancing the state-of-the-art in semantic and diversity metrics.

Findings

Semantic losses increased caption diversity from 0.07 to 0.18.

Using multiple visual features improved normalized captioning scores by 28%.

The model effectively integrates language and vision for detailed soccer video descriptions.

Abstract

This work aims at generating captions for soccer videos using deep learning. In this context, this paper introduces a dataset, model, and triple-level evaluation. The dataset consists of 22k caption-clip pairs and three visual features (images, optical flow, inpainting) for ~500 hours of \emph{SoccerNet} videos. The model is divided into three parts: a transformer learns language, ConvNets learn vision, and a fusion of linguistic and visual features generates captions. The paper suggests evaluating generated captions at three levels: syntax (the commonly used evaluation metrics such as BLEU-score and CIDEr), meaning (the quality of descriptions for a domain expert), and corpus (the diversity of generated captions). The paper shows that the diversity of generated captions has improved (from 0.07 reaching 0.18) with semantics-related losses that prioritize selected words. Semantics-related losses and the utilization of more visual features (optical flow, inpainting) improved the normalized captioning score by 28\%. The web page of this work: https://sites.google.com/view/soccercaptioning}{https://sites.google.com/view/soccercaptioning