一步一步地预见未来
Envisioning the Future, One Step at a Time
把未来场景演化预测从稠密视频/潜空间生成改为稀疏点轨迹上的逐步自回归扩散推理，使单张图像快速 rollout 上千种未来假设；对做视频预测、轨迹生成或世界模型的研究者，这是一种可复用的「以稀疏轨迹为生成载体」的建模范式，并配套开放集运动预测基准 OWM。
原文依据 formulating the prediction of open-set future scene dynamics as step-wise inference over sparse point trajectories
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论要准确预判复杂多样场景将如何演化，模型需要能够表征不确定性、沿长程交互链进行模拟，并高效探索大量可能的未来。然而现有方法大多依赖稠密视频或潜空间预测，把大量容量花在稠密外观上，而非场景中稀疏点的底层轨迹，这使得对未来假设的大规模探索代价高昂，并在长时程、多模态运动至关重要时限制了性能。为此，我们将开放集未来场景动态的预测建模为对稀疏点轨迹的逐步推理。我们的自回归扩散模型通过短程、局部可预测的转移来推进这些轨迹，并显式建模不确定性随时间的增长。这种以动态为中心的表征使得从单张图像快速 rollout 数千种多样的未来成为可能，并可选择由初始运动约束引导，同时保持物理合理性与长程连贯性。我们进一步提出 OWM，一个基于多样真实场景视频的开放集运动预测基准，用于在真实世界不确定性下评估预测轨迹分布的准确性与多样性。我们的方法在预测准确性上匹配或超越稠密模拟器，同时实现数量级更高的采样速度，使开放集未来预测既具可扩展性又切实可用。
英文原文摘要 1476 字符
Accurately anticipating how complex, diverse scenes will evolve requires models that represent uncertainty, simulate along extended interaction chains, and efficiently explore many plausible futures. Yet most existing approaches rely on dense video or latent-space prediction, expending substantial capacity on dense appearance rather than on the underlying sparse trajectories of points in the scene. This makes large-scale exploration of future hypotheses costly and limits performance when long-horizon, multi-modal motion is essential. We address this by formulating the prediction of open-set future scene dynamics as step-wise inference over sparse point trajectories. Our autoregressive diffusion model advances these trajectories through short, locally predictable transitions, explicitly modeling the growth of uncertainty over time. This dynamics-centric representation enables fast rollout of thousands of diverse futures from a single image, optionally guided by initial constraints on motion, while maintaining physical plausibility and long-range coherence. We further introduce OWM, a benchmark for open-set motion prediction based on diverse in-the-wild videos, to evaluate accuracy and variability of predicted trajectory distributions under real-world uncertainty. Our method matches or surpasses dense simulators in predictive accuracy while achieving orders-of-magnitude higher sampling speed, making open-set future prediction both scalable and practical.