Research
Publications and applied research outputs spanning action spotting, tactical classification, and real-time sports AI systems.
Featured Research
-
Proceedings Paper
From Raw Feeds to Curated Clips: A Modular AI Framework for Sports Highlight Production
Reduced post-event highlight processing time by 87.5% while maintaining 90% event-selection accuracy.
Read publication -
Conference Paper
Real-time classification of handball game situations
98.6% classification accuracy with 4 ms inference and 34.04 ms end-to-end pipeline latency.
Read publication
All Research
-
From Raw Feeds to Curated Clips: A Modular AI Framework for Sports Highlight ProductionReduced post-event highlight processing time by 87.5% while maintaining 90% event-selection accuracy.
Read publication -
Large Language Model Based Chatbot for Database Interaction through Natural LanguageDemonstrated practical NL-to-SQL interaction quality across four LLMs using human evaluation metrics.
Read publication -
Play by Play, a dataset of handball and basketball game situations in a standardized spaceReleased a synthetic labeled dataset with 365,383 sports frames covering seven standardized game situations.
Read publication -
OSL-ActionSpotting, A unified library for action spotting in sports videosUnified fragmented action-spotting methods into one reusable library pipeline for faster experimentation.
Read publication -
Beyond the Premier, Assessing action spotting transfer capability across diverse domainsBroadcast editing style appears as the dominant factor affecting cross-domain action-spotting transfer.
Read publication -
Real-time classification of handball game situations98.6% classification accuracy with 4 ms inference and 34.04 ms end-to-end pipeline latency.
Read publication