PlayGen-MoG:基于高斯混合轨迹预测的多智能体多样化play生成框架
PlayGen-MoG: Framework for Diverse Multi-Agent Play Generation via Mixture-of-Gaussians Trajectory Prediction
AI总结:
本文提出PlayGen-MoG框架,通过高斯混合输出头、相对空间注意力和非自回归预测方法,解决多智能体轨迹生成中的多样性与空间协调问题,实现从单一静态阵型生成真实play。
摘要:
多智能体轨迹生成在团队运动中需要能够捕捉可能play多样性和球员间真实空间协调的模型。标准生成方法如条件变分自编码器(CVAE)和扩散模型在这一任务上表现不佳,出现后验崩溃或收敛到数据集均值的问题。此外,大多数轨迹预测方法在预测模式下需要多个观察历史帧,限制了play设计中仅可用初始阵型的应用。我们提出了PlayGen-MoG,一个可扩展的阵型条件play生成框架,通过三个设计选择解决这些问题:1/ 一个具有共享混合权重的高斯混合(MoG)输出头,其中一组权重选择play场景,耦合所有球员的轨迹;2/ 相对空间注意力,将成对球员位置和距离编码为学习的注意力偏差;3/ 非自回归预测绝对位移,从初始阵型预测绝对位移,消除累积误差漂移并去除对观察轨迹历史的依赖,使从单一静态阵型生成真实play成为可能。在美式足球跟踪数据上,PlayGen-MoG实现了1.68码ADE和3.98码FDE,同时保持了全部8个混合组件的完全利用,熵值为2.06(最大2.08),并定性确认了多样化生成而无模式崩溃。
英文摘要:
Multi-agent trajectory generation in team sports requires models that capture both the diversity of possible plays and realistic spatial coordination between players on plays. Standard generative approaches such as Conditional Variational Autoencoders (CVAE) and diffusion models struggle with this task, exhibiting posterior collapse or convergence to the dataset mean. Moreover, most trajectory prediction methods operate in a forecasting regime that requires multiple frames of observed history, limiting their use for play design where only the initial formation is available. We present PlayGen-MoG, an extensible framework for formation-conditioned play generation that addresses these challenges through three design choices: 1/ a Mixture-of-Gaussians (MoG) output head with shared mixture weights across all agents, where a single set of weights selects a play scenario that couples all players' trajectories, 2/ relative spatial attention that encodes pairwise player positions and distances as learned attention biases, and 3/ non-autoregressive prediction of absolute displacements from the initial formation, eliminating cumulative error drift and removing the dependence on observed trajectory history, enabling realistic play generation from a single static formation alone. On American football tracking data, PlayGen-MoG achieves 1.68 yard ADE and 3.98 yard FDE while maintaining full utilization of all 8 mixture components with entropy of 2.06 out of 2.08, and qualitatively confirming diverse generation without mode collapse.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。