FootBots: A Transformer-based Architecture for Motion Prediction in Soccer
This addresses motion prediction for soccer analytics, but it appears incremental as it builds on existing transformer methods with specific adaptations.
The paper tackles motion prediction in soccer by capturing complex player and ball interactions, presenting FootBots, a transformer-based architecture that outperforms baselines in motion prediction and excels in conditioned tasks like predicting players based on ball position.
Motion prediction in soccer involves capturing complex dynamics from player and ball interactions. We present FootBots, an encoder-decoder transformer-based architecture addressing motion prediction and conditioned motion prediction through equivariance properties. FootBots captures temporal and social dynamics using set attention blocks and multi-attention block decoder. Our evaluation utilizes two datasets: a real soccer dataset and a tailored synthetic one. Insights from the synthetic dataset highlight the effectiveness of FootBots' social attention mechanism and the significance of conditioned motion prediction. Empirical results on real soccer data demonstrate that FootBots outperforms baselines in motion prediction and excels in conditioned tasks, such as predicting the players based on the ball position, predicting the offensive (defensive) team based on the ball and the defensive (offensive) team, and predicting the ball position based on all players. Our evaluation connects quantitative and qualitative findings. https://youtu.be/9kaEkfzG3L8