Authoring Narrative Visualization in Motion: Visual Storytelling in Swimming Videos

A new HCI/graphics project tackles a familiar production problem: how to turn dense sports footage into a clear, data-driven story without bolting on static overlays after the fact. The team focused on swimming races because they combine fast motion, lane-based comparisons, and timing data that can be visualized in ways viewers immediately understand.

The core of the work is an automated pipeline that pulls structured information from video, then pairs it with narrative constructs identified through observational analysis of sports broadcasts. That data is then used in a technology probe designed to help creators author visualizations that move with the race and coordinate views, data, and temporal progression.

For game developers, the interesting part is less the sport itself and more the production pattern: this is a concrete example of blending computer vision, data extraction, and authoring tools into a single storytelling workflow. The same ideas map to replay systems, esports broadcasts, training tools, and any game-adjacent video where timing, position, and progression need to be explained clearly.

The team evaluated the approach with experienced content creators and graphic designers to surface both benefits and friction points. That makes the work relevant to tool builders and technical artists who are trying to reduce manual motion-graphics labor while keeping enough editorial control for storytelling.