MeshUp #7 – Pål Halvorsen
Pål Halvorsen will present a short lecture on the topic From Broadcast to Feed: AI-Driven Sports Highlight Production at Scale.
Abstract
Sports organizations and broadcasters face growing pressure to feed an always-on content cycle across multiple platforms, yet highlight production remains a largely manual, resource-intensive process. Failing to publish timely content means missed opportunities to reach fans when it matters most, directly impacting audience growth, sponsorship value, and brand loyalty. However, advances in computer vision, natural language processing, and generative AI now make it possible to automate tasks that once demanded significant human expertise — from recognizing key moments in a broadcast feed to crafting platform-tailored captions. For sports media, where content volume is high and the window for relevance is short, these tools represent a genuine step change in productivity. This talk presents an AI-based multimedia production framework that addresses this challenge directly, enabling the automatic generation and publishing of soccer and ice hockey highlights at scale — hopefully without sacrificing the quality fans demand.
Bio
Pål Halvorsen is a research scientist at SimulaMet, a professor at OsloMet and at the University of Oslo. His interests include complete end-to-end pipelines with a particular focus on sport and medical applications. In this respect, his current research interests span several areas in distributed systems and content analysis from both a performance and efficiency point of view. He is also co-founder and CEO of Forzasys, the company from which the work presented in this talk originates.
Access
MishMash MeetUps are short, informal meetings in the consortium where both early career and established researchers present ongoing projects. The events are open for everyone, but, for security reasons, Zoom links are only provided to people that are affiliated with a MishMash Work Package. If not, please ask for access.