R&D - Modeling Key World Cup Moments with Machine Learning

A Times journalist on location in Qatar photographs the match, capturing images in high-speed bursts, trying to anticipate important moments with a broad view of the field. In collaboration with The Times's newsroom in the U.S., they identify a key moment in the match and transmit the photograph of that moment. From that photo, we calculate the 3D position of the photographer's camera. We can do so mathematically using the standard size of the goal and penalty area as geometric guides. Once we know the camera position, we project the image onto 3D geometry that represents the field and stands where the game was played.