US Patent 11369885 Telemetric video processing

Various aspects of the subject technology relate to systems, methods, and machine-readable media for extracting information from videos. The method includes annotating portions of interest within screen captures. The method also includes receiving at least a first set of videos for a video game. The method also includes training a first machine-learning model to identify the portions of interest within the first set of videos. The method also includes generating validation data based on results of the first machine-learning model. The method also includes extracting information based on the portions of interest identified in the first set of videos.