Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces FlowS, a novel framework for one-step motion prediction that leverages local transport conditioning to achieve high accuracy and diverse multimodal futures while maintaining low latency. By conditioning the base distribution to focus on plausible futures, the model reduces the complexity of motion prediction to short-range refinements, enabling accurate predictions in a single step. On the Waymo Open Motion Dataset, FlowS outperforms existing methods, achieving state-of-the-art metrics with single-step inference at 75 FPS, highlighting its potential for real-world autonomous applications.
Local transport conditioning allows FlowS to achieve state-of-the-art motion prediction accuracy in a single inference step, making it a game-changer for safety-critical autonomous systems.
Generative motion prediction must satisfy three simultaneous requirements for real-world autonomy: high accuracy, diverse multimodal futures, and strictly bounded latency. Diffusion models meet the first two but violate the third, requiring tens to hundreds of denoising steps. We identify a conditioning strategy that resolves this tension: \textit{single-step integration is accurate when the underlying transport problem is local}. A model that must both discover the correct behavioral mode and traverse a long displacement in one step accumulates large discretization errors; conditioning the base distribution to lie near plausible futures reduces the problem to short-range refinement, the regime where a single Euler step suffices. We instantiate this \emph{local transport conditioning} in FlowS, a conditional flow matching framework with two mechanisms. First, an online, scene-conditioned learned prior emits $K$ calibrated anchor trajectories per agent, each already near a plausible future, converting mode discovery into local correction. Second, a step-consistent displacement field enforces semigroup self-consistency, guaranteeing that a single step inherits multi-step accuracy. Crucially, anchoring this field at learned priors along straight-line paths yields a {stable, low-variance} training target, unlike prior self-consistency methods that suffer from {high-variance bootstrap} signals on curved diffusion paths. On the Waymo Open Motion Dataset, FlowS achieves state-of-the-art Soft mAP {(0.4804) and mAP (0.4703) with ensemble at 75\,FPS} with single-step inference, demonstrating that local transport conditioning makes one-step generative motion prediction practical for safety-critical autonomy. Code and pretrained models will be released upon acceptance.