Semantic Action Graph: A Shared Representation for Agent Grounding and Human Interpretation of Sports Highlights
AI Digest
语义动作图代理生成人类解释体育视频分析结构化表示
文章提出语义动作图，通过结构化表示解决AI生成体育视频亮点的可解释性问题，同时支持人类查询与分析。
This paper introduces a semantic action graph to address AI-generated sports highlights' interpretability, enabling both agent grounding and human analysis through structured representation.
Key points
- 提出语义动作图作为轻量级领域模式，整合表演者、动作、时刻等要素 Introduces semantic action graph as a lightweight domain schema integrating performers, actions, and moments
- 三重核心属性：事件序列连接、共享词汇、可定位帧级时刻 Three core properties: connected event sequences, shared vocabulary, and frame-addressable moments
- SportSAGE系统验证显示用户可有效通过图界面查询视频亮点 SportSAGE system validation shows users can effectively query highlights via graph interface
- 解决生成式代理输出难以验证与个性化调整的行业痛点 Addresses the industry challenge of verifying and personalizing AI-generated content
Takeaway: 结构化语义框架可同时满足AI生成与人类交互需求 / Structured semantic framework satisfies both AI generation and human interaction needs
Why it matters 提供AI与用户交互的新范式，对体育内容生成有实践指导价值
View original ↗ Back to hot list
This page is an aggregated digest from arxiv; content and hot-score data come from public sources. Copyright belongs to the original authors. We link to originals with nofollow and never republish full text.