Capturing Uncertainty in Human Motion for Representation Learning in Soccer
2026-08-11 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors developed a method for teaching computers to understand soccer players' 3D movements by predicting what they will do next. They focus on capturing many possible future actions because human movement isn't always predictable. Their model learns this by looking at actual future movements as guidance. Tests show their method predicts motion more accurately and the learned knowledge can be used in other soccer-related tasks.
self-supervised learning3D skeleton-based motionfuture motion predictionprobabilistic distributionmultimodalitytrajectorymotion dynamicsrepresentation learningsoccer player tracking
Authors
Yizhou Xu, Lars Bretzner, Tiesheng Wang, Atsuto Maki
Abstract
This paper presents a self-supervised representation learning framework for understanding 3D skeleton-based human motion in soccer, using future motion prediction as the learning objective. Since human motion is inherently uncertain, accounting for multiple plausible futures is essential for capturing the underlying motion dynamics and learning effective representations. To this end, we introduce a conditioning module for motion prediction that models a probabilistic distribution over discretized future motions in 3D Euclidean space, learning multimodality with explicit supervision from future trajectories. Experiments on large-scale soccer player tracking data show that our approach substantially improves motion prediction accuracy. Moreover, the learned representations effectively transfer to multiple soccer downstream applications, demonstrating strong cross-task generalization.