arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2610.37583cs.CVcs.LGcs.MM
收藏

足球转播中的零样本多模态事件检测

Zero-Shot Multimodal Event Spotting in Soccer Broadcasts

University of Maryland Baltimore County(马里兰大学巴尔的摩县分校);Comcast Applied AI(康卡斯特应用人工智能);Comcast(康卡斯特)

Sadia Kamal, Mahmudul Hasan, Tony Braskich

AI总结:

本文提出零样本多模态框架,无需任务特定训练即可检测足球转播中的事件,在3场含130个事件的比赛上优于基准模型,推理成本降低72%,还引入兴奋度评分阶段用于精彩集锦选择。

摘要:

自动足球事件检测可用于生成精彩集锦、视频索引和体育分析,但现有方法依赖大型标注数据集和针对足球的特定训练。本文提出一种零样本框架,无需任务特定参数训练即可检测、分类并时间定位完整足球转播中的事件。每场比赛被分割为重叠片段,由紧凑的视觉故事板和对齐的解说文本表示,通用多模态模型对二者进行联合处理。在包含130个事件的3个人工标注比赛上,最优配置达到72.0%的F1@5s指标,在跨数据集设置中优于经SoccerNet训练的CALF和NetVLAD++;与单帧相比,故事板表示将推理成本降低72%。本文进一步分析了模型家族、输入表示和提示优化的影响,并引入基于证据的兴奋度评分阶段,该阶段为检测到的事件排序并提供自然语言解释以用于精彩集锦选择。

英文摘要:

Automatic soccer event spotting supports highlight generation, video indexing, and sports analytics, but existing methods depend on large annotated datasets and soccer-specific training. We present a zero-shot framework that detects, classifies, and temporally localizes events in full-length soccer broadcasts without task-specific parameter training. Each match is segmented into overlapping clips represented by a compact visual storyboard and the aligned commentary transcript, which a general-purpose multimodal model processes jointly. On three manually annotated matches containing 130 events, the best configuration achieves 72.0% F1@5s and outperforms SoccerNet-trained CALF and NetVLAD++ in a cross-dataset setting, while the storyboard representation reduces inference cost by 72% compared with individual frames. We further analyze the effects of model family, input representation, and prompt optimization, and introduce an evidence-grounded excitement-scoring stage that ranks detected events with natural-language explanations for highlight selection.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑