Multi-Modal Trajectory Prediction of NBA Players

Hauri, Sandro; Djuric, Nemanja; Radosavljević, Vladan; Vučetić, Slobodan

doi:10.1109/wacv48630.2021.00168

Cited by 15 publications

(15 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the ELBO (15), we anneal β from an initial value of 0.1 to final values of 0.01 and 1 in our sweeps. For all bidirectional models, we sweep over the two fusion modes specified in (13) and (14). For each model, training for each hyperparameter set is conducted and reported over a sweep of 5 random seeds.…”

Section: Discussionmentioning

confidence: 99%

“…There exist a number of works from various fields of research that are related to our approach. Specifically, a number of works from robotics and computer vision [1,25,26,40], sports analytics [3,13,17,32,38], economics [28,35,36] and machine learning [6,11,33] focus on various combinations of missing data imputation and multiagent trajectory predictions. Given the broad scope of time series prediction as a research field [12], we focus particularly on models that predict human trajectories [25], as they are the most relevant for our problem regime.…”

Section: Related Workmentioning

confidence: 99%

“…Considers interactions between agents Imputation Forward & backward inf MBT [13] Basketball VAIN [15] No seq. pred.…”

Section: Stochasticmentioning

confidence: 99%

“…Bidir. fusion mode: the fusion mode used for bidirectional models, where 'mean' corresponds to (13) in the main text, and 'nearest' to (14). For each baseline model, we compute the mean evaluation loss, L 2 (Mean), compared to the ground truth trajectories (over all seeds).…”

Section: Supplementary Material: Time-series Imputation Of Temporally...mentioning

confidence: 99%

“…Predictive modeling of multiagent behaviors has been a topic of considerable interest in machine learning [6,11,33], financial economics [28,35,36], robotics [1,25,26], and sports analytics [13,17,18,38]. In such systems, decision-making agents interact within a shared environment, following an underlying dynamical process that may be stochastic and, at times, infeasible to characterize analytically due to the complex interactions involved.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Time-series Imputation of Temporally-occluded Multiagent Trajectories

Omidshafiei,

Hennes,

Garnelo

et al. 2021

Preprint

View full text Add to dashboard Cite

In multiagent environments, several decision-making individuals interact while adhering to the dynamics constraints imposed by the environment. These interactions, combined with the potential stochasticity of the agents' decision-making processes, make such systems complex and interesting to study from a dynamical perspective. Significant research has been conducted on learning models for forward-direction estimation of agent behaviors, for example, pedestrian predictions used for collision-avoidance in self-driving cars. However, in many settings, only sporadic observations of agents may be available in a given trajectory sequence. For instance, in football, subsets of players may come in and out of view of broadcast video footage, while unobserved players continue to interact off-screen. In this paper, we study the problem of multiagent time-series imputation, where available past and future observations of subsets of agents are used to estimate missing observations for other agents. Our approach, called the Graph Imputer, uses forward-and backward-information in combination with graph networks and variational autoencoders to enable learning of a distribution of imputed trajectories. We evaluate our approach on a dataset of football matches, using a projective camera module to train and evaluate our model for the off-screen player state estimation setting. We illustrate that our method outperforms several state-of-the-art approaches, including those hand-crafted for football.Preprint. Under review.

show abstract

Section: Discussionmentioning

confidence: 99%