发表机构
Syracuse University（雪城大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对非周期视觉-运动时间序列，提出利用内外生变量的深度学习架构及i.i.d采样训练，在跌倒预测任务上显著超越现有方法。
AI 中文摘要
近年来，深度学习模型越来越多地被应用于时间序列预测（TSF）。基于Transformer和基于MLP的模型都已在许多真实世界的TSF回归基准上得到有效应用，并且关于哪一类方法更优的争论仍在持续。尽管这些基准引起了广泛关注，但同样值得注意的是，许多当前的数据集和方法假设时间序列具有近似周期性。在本工作中，我们关注一个无周期性的新TSF任务：基于第一人称视觉和本体感觉来预测人形机器人运动过程中的跌倒。当运动轨迹足够多样化时，周期性即被破坏。我们贡献了两个新的基准数据集（一个来自仿真，一个来自真实硬件），表明周期性在这些基准上确实被破坏，且近期的深度TSF方法在这些基准上表现不佳。我们还提出了一种新颖的深度学习架构，该架构同时利用内生变量和外生变量，并采用一种严格强制训练样本独立同分布（i.i.d）采样的训练过程。我们的结果显示，在多种实验条件下，相较于先前最先进的方法，在真实数据上实现了至少12.73%的统计显著提升，在仿真数据上实现了至少10.40%的提升。代码和数据集将在论文被接收后提供。
英文摘要
Deep learning models have been increasingly applied to Time Series Forecasting (TSF) in recent years. Transformer-based and MLP-based models have both been used effectively on many real-world TSF regression benchmarks, and there is ongoing debate as to which family of methods is best. While these benchmarks have drawn much attention, it is also worth noting that many current datasets and methods assume approximate periodicity in the time series. In this work, we focus on a new TSF task without periodicity: anticipating falls during humanoid locomotion, on the basis of egocentric vision and proprioception. When the locomotion trajectories are sufficiently diverse, periodicity is violated. We contribute two new benchmark datasets (one from simulation, one from real hardware), showing that periodicity is violated and recent deep TSF methods struggle on these benchmarks. We also propose a novel deep learning architecture that exploits both endogenous and exogenous variables and a training process that rigorously enforces i.i.d sampling of training examples. Our results show statistically significant improvement over prior art in multiple experimental conditions, by 12.73% or more on the real data and 10.40% or more on the simulation data. Code and datasets will be available upon acceptance.
