Authoring Narrative Visualization in Motion: Visual Storytelling in Swimming Videos

2026-07-16

AI Summary

Existing approaches struggle to effectively integrate perspective, data, and temporal dimensions to convey compelling narratives in sports videos. This work proposes the first narrative visualization authoring framework specifically designed for sports video storytelling, using competitive swimming as a case study. The framework automatically extracts structured data from videos, analyzes narrative patterns commonly employed in sports broadcasting, and provides a dynamic visualization authoring tool. It enables content creators to synthesize multidimensional information into coherent and expressive visual narratives of athletic events. Expert evaluations demonstrate that the proposed method significantly enhances both the comprehensibility of complex spatiotemporal dynamics and the overall narrative expressiveness in sports storytelling.

Abstract

We investigate how to support authoring narrative visualizations in motion in sports videos, drawing on automated data preparation, systematic analysis, technology probe design, and evaluation, using swimming races as a case study. Sports videos are widely broadcast and shared across social media, where content creators increasingly seek to present and explain complex events to general audiences. Visualization in motion has been explored as an efficient way to embed data into videos and to move with the data referents, providing additional information and helping audiences understand races. However, existing approaches primarily focus on embedding visualizations in videos, lacking exploration of how to support authoring narratives that coordinate views, data, and temporal progression to explain the unfolding races. To address this gap, we use swimming videos as an ideal case for exploration, as swimming is a sport with rich, dynamic data and visualizations in practice. We develop an automated pipeline that extracts structured data from videos, derive narrative constructs through observational analysis of sports broadcasts, and design a technology probe that supports authoring using data prepared by our pipeline and narrative constructs derived from our observations. We evaluate our approach with experienced content creators and/or graphic designers to examine the benefits and challenges of authoring narrative visualizations in motion. All supplemental materials are described in the Supplemental Material Pointers section and are on OSF: osf.io/bq47n/.

Problem

• narrative visualization
• sports videos
• visual storytelling
• authoring tools
• temporal progression

Innovation

• narrative visualization
• visualization in motion
• automated data extraction
• sports video analysis
• authoring tool