Paper / Slides / Poster / Code / Video

As artificial intelligence spreads out to numerous fields, the application of AI to sports analytics is also in the spotlight. However, one of the major challenges is the difficulty of automated acquisition of continuous movement data during sports matches. In particular, it is a conundrum to reliably track a tiny ball on a wide soccer pitch with obstacles such as occlusion and imitations. Tackling the problem, this study proposes an inference framework of ball trajectory from player trajectories as a cost-efficient alternative to ball tracking. We combine Set Transformers to get permutation-invariant and equivariant representations of the multi-agent contexts with a hierarchical architecture that intermediately predicts the player ball possession to support the final trajectory inference. Also, we introduce the reality loss term and postprocessing to secure the estimated trajectories to be physically realistic. The experimental results show that our model provides natural and accurate trajectories as well as admissible player ball possession at the same time. Lastly, we suggest several practical applications of our framework including missing trajectory imputation, semi-automated pass annotation, automated zoom-in for match broadcasting, and calculating possession-wise running performance metrics.

Paper / Slides / Poster / Code / Video

In fluid team sports such as soccer and basketball, analyzing team formation is one of the most intuitive ways to understand tactics from domain participants' point of view. However, existing approaches either assume that team formation is consistent throughout a match or assign formations frame-by-frame, which disagree with real situations. To tackle this issue, we propose a change-point detection framework named SoccerCPD that distinguishes tactically intended formation and role changes from temporary changes in soccer matches. We first assign roles to players frame-by-frame and perform two-step change-point detections: (1) formation change-point detection (FormCPD) based on the sequence of role-adjacency matrices and (2) role change-point detection (RoleCPD) based on the sequence of role permutations. The evaluation of SoccerCPD using the ground truth annotated by domain experts shows that our method accurately detects the points of tactical changes and estimates the formation and role assignment per segment. Lastly, we introduce practical use-cases that domain participants can easily interpret and utilize.

Paper / Slides

Although the values of individual soccer players have become astronomical, subjective judgments still play a big part in the player analysis. Recently, there have been new attempts to quantitatively grasp players' styles using video-based event stream data. However, they have some limitations in scalability due to high annotation costs and sparsity of event stream data. In this study, we build a triplet network named 6MapNet that can effectively capture the movement styles of players using in-game GPS data. Without any annotation of soccer-specific actions, we use players' locations and velocities to generate two types of heatmaps. Our subnetworks then map these heatmap pairs into feature vectors whose similarity corresponds to the actual similarity of playing styles. The experimental results show that players can be accurately identified with only a small number of matches by our method.

Paper / Slides / Poster / FIFA Certificate

Player tracking data are now widely used in the sports industry to provide deeper insights to domain participants. Global positioning systems (GPS) and camera-based optical tracking systems (OTS) are two common tracking systems, but the former suffers from location biases, and the latter requires a heavy installment of multiple cameras or a lot of manual correction work. In this study, we propose a framework for cost-efficient and bias-robust player tracking by integrating GPS and video data. We design a sophisticated filtering algorithm to selectively use the positional information from bounding boxes detected in the video and use the GPS data as a reliable tool for identifying the chosen boxes. Using the player identity and location information of these bounding boxes, we estimate and remove GPS biases in two steps to obtain unbiased player trajectories. We demonstrate that our algorithm precisely tracks players from video with the aid of GPS data even in poor conditions such as the presence of player occlusions and players outside the sight of cameras.