BoxComm: Benchmarking Category-Aware Commentary Generation and Narration Rhythm in Boxing

2026-04-06

AI Summary

Existing research on sports commentary generation has predominantly focused on team sports, overlooking the unique challenges posed by combat sports such as boxing—characterized by highly transient actions, subtle yet semantically critical visual distinctions, and dense tactical analysis. To address this gap, this work introduces BoxComm, a large-scale boxing commentary dataset comprising 445 match videos and 52,000 professional commentaries, featuring the first-ever category-level annotations. The authors propose a novel commentary taxonomy encompassing live play-by-play, tactical analysis, and contextual background, and establish two new evaluation tasks: category-conditioned generation and commentary pacing assessment. Building upon multimodal large language models and structured action prompts derived from punch event detection, they develop EIC-Gen, an enhanced baseline model. Experiments demonstrate that existing models perform poorly on these new tasks, whereas EIC-Gen significantly improves generation quality, underscoring the critical role of fine-grained action awareness in combat sports commentary generation.

Abstract

Recent multimodal large language models (MLLMs) have shown strong capabilities in general video understanding, driving growing interest in automatic sports commentary generation. However, existing benchmarks for this task focus exclusively on team sports such as soccer and basketball, leaving combat sports entirely unexplored. Notably, combat sports present distinct challenges: critical actions unfold within milliseconds with visually subtle yet semantically decisive differences, and professional commentary contains a substantially higher proportion of tactical analysis compared to team sports. In this paper, we present BoxComm, a large-scale dataset comprising 445 World Boxing Championship match videos with over 52K commentary sentences from professional broadcasts. We propose a structured commentary taxonomy that categorizes each sentence into play-by-play, tactical, or contextual, providing the first category-level annotation for sports commentary benchmarks. Building on this taxonomy, we introduce two novel and complementary evaluations tailored to sports commentary generation: (1) category-conditioned generation, which evaluates whether models can produce accurate commentary of a specified type given video context; and (2) commentary rhythm assessment, which measures whether freely generated commentary exhibits appropriate temporal pacing and type distribution over continuous video segments, capturing a dimension of commentary competence that prior benchmarks have not addressed. Experiments on multiple state-of-the-art MLLMs reveal that current models struggle on both evaluations. We further propose EIC-Gen, an improved baseline incorporating detected punch events to supply structured action cues, yielding consistent gains and highlighting the importance of perceiving fleeting and subtle events for combat sports commentary.

Problem

• sports commentary generation
• combat sports
• commentary rhythm
• category-aware evaluation
• boxing

Innovation

• category-aware commentary
• commentary rhythm
• combat sports
• structured action cues
• multimodal benchmark