Summary
Figuring out where players are on a soccer field from TV footage is hard because you want their exact position on the real field, not just how their body moves. The authors created Field Converter, a method that starts by using the known camera angles and field layout to guess where a player’s feet touch the ground. Then, it makes small adjustments by looking at video frames over time and combining different clues. This approach makes the player position much more accurate, cutting error by more than half compared to just using geometry. They found that using clues from multiple frames helps more than the choice of technical details, but the method struggles a bit when players jump or are airborne.
3D human pose estimationmonocular sports broadcastworld coordinate systemcamera calibrationgeometry initializationtemporal residual correctionMLPtemporal convolutional networkTransformermean per joint position error (MPJPE)
Authors
Simon Khan, Laurent Gajny, Jennyfer Lecompte, Sébastien Laporte
Abstract
Recovering 3D human pose from monocular sports broadcasts remains challenging when players must be localized in a shared metric world coordinate system rather than only reconstructed relative to their own body. We introduce Field Converter, a geometry-initialized temporal residual framework for world-grounded 3D player pose estimation from calibrated soccer broadcasts. Our method first uses camera and pitch geometry to initialize the player root through ray-ground intersection, then predicts a temporal residual correction from pose, image, camera, and geometric cues. On match-disjoint evaluation sequences, residual refinement reduces root error from 49cm with geometry alone to 14cm with a frame-wise MLP and 10cm with a TCN, while a Transformer achieves a comparable 11cm. The resulting world-space MPJPE reaches 13.2cm, and ablations show that residual prediction clearly outperforms direct global-root regression while temporal context matters more than the specific temporal backbone. Failure analysis further identifies airborne motion as the main limitation of the ground-based geometric initialization.