Envisioning the Future, One Step at a Time

• URL: http://arxiv.org/abs/2604.09527v1
• Date: Fri, 10 Apr 2026 17:46:05 GMT
• Title: Envisioning the Future, One Step at a Time
• Abstract summary: Open-set future scene dynamics is predicted as step-wise inference over sparse point trajectories.<n>Our autoregressive diffusion model advances these trajectories through short, locally predictable transitions.<n>We introduce OWM, a benchmark for open-set motion prediction based on diverse in-the-wild videos.
• Score: 22.38380614056375
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: 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. Project page: http://compvis.github.io/myriad.

Related papers

• Drive-HWM: Hierarchical World Models for Dynamic-Latent Guided Autonomous Driving [78.83165764484562]
  Drive-HWM is a hierarchical slow--fast world modeling framework that organizes future representation prediction and action generation at complementary temporal scales.<n>Next-frame prediction encourages the fast model to capture imminent scene evolution, while one-step action generation allows decisions to be continuously updated as new observations arrive.
  arXiv  Detail & Related papers  (2026-09-03T09:21:19Z)
• Schrödinger's Cat: Probabilistic Representation and Prediction of Potential Scene Kinematics [18.394277994509956]
  Predicting how a scene may evolve from partial observations requires reasoning about multiple possible futures rather than committing to a single trajectory.<n>We introduce Goal-Aware Representations of Future kEmatic Latent Distributions (GAR)<n>GAR learns structured sparse-temporal latent representation of distribution over possible futures given an image and optional-temporally sparse constraints.<n>As a result, uncertainty about future motion can be localized to specific scene elements and timesteps and progressively refined through additional constraints.
  arXiv  Detail & Related papers  (2026-07-28T17:05:39Z)
• SegDiff: Segmented Trajectory Diffusion for Consistent and Adaptive Robot Manipulation [65.81476183180527]
  We introduce SegDiff, a closed-loop visuomotor policy that integrates the strengths of both paradigms.<n>SegDiff decomposes demonstrations into motion segments between keyposes and learns to predict the continuous trajectory from the current state to the next keypose.<n>SegDiff demonstrates significant performance gains over existing approaches across various simulated and real-world scenarios.
  arXiv  Detail & Related papers  (2026-07-13T02:48:03Z)
• Gated Temporal Diffusion for Stochastic Long-Term Dense Anticipation [17.4088244981231]
  Long-term action anticipation has become an important task for many applications such as autonomous driving and human-robot interaction. We propose a novel Gated Temporal Diffusion (GTD) network that models the uncertainty of both the observation and the future predictions. Our model achieves state-of-the-art results on the Breakfast, Assembly101 and 50Salads datasets in both deterministic settings.
  arXiv  Detail & Related papers  (2024-07-16T17:48:05Z)
• State-space Decomposition Model for Video Prediction Considering Long-term Motion Trend [3.910356300831074]
  We propose a state-space decomposition video prediction model that decomposes the overall video frame generation into deterministic appearance prediction and motion prediction. We infer the long-term motion trend from conditional frames to guide the generation of future frames that exhibit high consistency with the conditional frames.
  arXiv  Detail & Related papers  (2024-04-17T17:19:48Z)
• PREF: Predictability Regularized Neural Motion Fields [68.60019434498703]
  Knowing 3D motions in a dynamic scene is essential to many vision applications. We leverage a neural motion field for estimating the motion of all points in a multiview setting. We propose to regularize the estimated motion to be predictable.
  arXiv  Detail & Related papers  (2022-09-21T22:32:37Z)
• Graph-based Spatial Transformer with Memory Replay for Multi-future Pedestrian Trajectory Prediction [13.466380808630188]
  We propose a model to forecast multiple paths based on a historical trajectory. Our method can exploit the spatial information as well as correct the temporally inconsistent trajectories. Our experiments show that the proposed model achieves state-of-the-art performance on multi-future prediction and competitive results for single-future prediction.
  arXiv  Detail & Related papers  (2022-06-12T10:25:12Z)
• Stochastic Trajectory Prediction via Motion Indeterminacy Diffusion [88.45326906116165]
  We present a new framework to formulate the trajectory prediction task as a reverse process of motion indeterminacy diffusion (MID) We encode the history behavior information and the social interactions as a state embedding and devise a Transformer-based diffusion model to capture the temporal dependencies of trajectories. Experiments on the human trajectory prediction benchmarks including the Stanford Drone and ETH/UCY datasets demonstrate the superiority of our method.
  arXiv  Detail & Related papers  (2022-03-25T16:59:08Z)
• StretchBEV: Stretching Future Instance Prediction Spatially and Temporally [0.0]
  In self-driving cars, predicting future in terms of location and motion of all the agents around the vehicle is a crucial requirement for planning. Recently, a new joint formulation of perception and prediction has emerged by fusing rich sensory information perceived from multiple cameras into a compact bird's-eye view representation to perform prediction. However, the quality of future predictions degrades over time while extending to longer time horizons due to multiple plausible predictions. In this work, we address this inherent uncertainty in future predictions with a temporal model.
  arXiv  Detail & Related papers  (2022-03-25T13:28:44Z)
• LookOut: Diverse Multi-Future Prediction and Planning for Self-Driving [139.33800431159446]
  LookOut is an approach to jointly perceive the environment and predict a diverse set of futures from sensor data. We show that our model demonstrates significantly more diverse and sample-efficient motion forecasting in a large-scale self-driving dataset.
  arXiv  Detail & Related papers  (2021-01-16T23:19:22Z)
• Long Term Motion Prediction Using Keyposes [122.22758311506588]
  We argue that, to achieve long term forecasting, predicting human pose at every time instant is unnecessary. We call such poses "keyposes", and approximate complex motions by linearly interpolating between subsequent keyposes. We show that learning the sequence of such keyposes allows us to predict very long term motion, up to 5 seconds in the future.
  arXiv  Detail & Related papers  (2020-12-08T20:45:51Z)
• From Goals, Waypoints & Paths To Long Term Human Trajectory Forecasting [54.273455592965355]
  Uncertainty in future trajectories stems from two sources: (a) sources known to the agent but unknown to the model, such as long term goals and (b)sources that are unknown to both the agent & the model, such as intent of other agents & irreducible randomness indecisions. We model the epistemic un-certainty through multimodality in long term goals and the aleatoric uncertainty through multimodality in waypoints& paths. To exemplify this dichotomy, we also propose a novel long term trajectory forecasting setting, with prediction horizons upto a minute, an order of magnitude longer than prior works.
  arXiv  Detail & Related papers  (2020-12-02T21:01:29Z)
• Temporally-Continuous Probabilistic Prediction using Polynomial Trajectory Parameterization [12.896275507449936]
  A commonly-used representation for motion prediction of actors is a sequence of waypoints for each actor at discrete future time-points. This approach is simple and flexible, but it can exhibit unrealistic higher-order derivatives and approximation errors at intermediate time steps. We propose a simple and general representation for temporally continuous trajectory prediction that is based on trajectory parameterization.
  arXiv  Detail & Related papers  (2020-11-01T01:51:44Z)
• SMART: Simultaneous Multi-Agent Recurrent Trajectory Prediction [72.37440317774556]
  We propose advances that address two key challenges in future trajectory prediction. multimodality in both training data and predictions and constant time inference regardless of number of agents.
  arXiv  Detail & Related papers  (2020-07-26T08:17:10Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.