LPA-CWM introduces a Learned Physical Adjudicator (LPA) to improve counterfactual world models (CWM) for motion reasoning by learning to weight candidate responses based on visual context and response structure. The 3.0M‑parameter LPA is trained on dense MOVi‑F trajectories while keeping the CWM predictor and intervention generator frozen. A new Completeness‑aware Motion Correspondence (CMC) protocol evaluates localization, trajectory completeness, visibility, and continuity, and LPA‑CWM achieves significant gains on DAVIS and Kinetics subsets.
arXiv:2610.03691v1 Announce Type: new Abstract: We present FlowHMR, a framework for recovering physically plausible global 3D human motion from monocular video. Previous learning-based methods typica...
We present FlowHMR, a framework for recovering physically plausible global 3D human motion from monocular video. Previous learning-based methods typically regress human motion directly from video and...
The paper introduces a reinforcement learning post‑training scheme that trains robot world models on their own autoregressive rollouts, using a contrastive RL objective adapted from diffusion models. It also proposes a training protocol that compares multiple variable‑length futures, a multi‑view visual fidelity reward, and demonstrates state‑of‑the‑art rollout fidelity on the DROID dataset, outperforming baselines on LPIPS, SSIM, and human preference tests.
arXiv:2610.26903v1 Announce Type: new Abstract: Action-conditioned world models support planning and policy evaluation by predicting future states. However, learning an accurate one-step predictor ca...
arXiv:2610.20604v1 Announce Type: new Abstract: World-action models aim to turn visual anticipation into driving decisions, yet typically train action experts on trajectory representations defined in...
WorldFlow is a generative world model that jointly predicts future video frames and query-based 3D trajectories using a unified Mixture-of-Transformers backbone. It represents 3D motion as continuous displacement tokens aligned with the video grid and optimizes a rectified-flow objective to couple explicit geometric dynamics with visual evolution. Trained on 1.44 million heterogeneous samples and evaluated on WorldFlowBench, it achieves the lowest average and final displacement errors among learning-based methods, notably outperforming baselines on spatial reasoning tasks.
The paper introduces Action Forcing, a method that transforms ordinary unlabeled video into action‑supervised training data by extracting egomotion bases through principal component analysis of pixel displacements. This approach yields grounded throttle–yaw control signals without requiring instrumented platforms or manual annotation, and it trains a high‑capacity video model while preventing pixel‑level overfitting via an online latent critic. The authors also critique standard video generation metrics and propose a reference‑free evaluation that measures controllability, plausibility, conjuring, and geometric integrity, showing that their model can reverse, scale, and compose actions despite limited reverse‑action data.
arXiv:2610.07922v1 Announce Type: cross Abstract: World-action models (WAMs) couple future prediction with robot control, yet existing systems often vary the video backbone, interaction structure, su...
The paper introduces HF‑WM, an action‑conditioned world model that learns from large‑scale egocentric human videos and incorporates failure dynamics through post‑training. It bridges the embodiment gap with a dual‑branch framework that couples video generation to stabilized pointflow prediction, unifying motion representation across humans and robots while reducing egocentric camera motion. Failure modeling is enhanced by Group Relative Policy Optimization on perturbed actions, guided by action alignment and physical plausibility rewards, enabling the model to anticipate unsuccessful executions. Experiments show HF‑WM outperforms existing methods in action‑conditioned video generation and is useful for evaluating policies in downstream robotic tasks.
SimVTA is a post‑training framework that enhances driving world action models by aligning simulated future videos with candidate trajectories. It samples trajectories, renders high‑reward futures in a real‑world simulator using 3D Gaussian Splatting, and prioritizes video supervision for high‑error futures. The approach improves planning performance, raising PDMS from 86.2 to 91.3 on NAVSIM v1 and achieving strong results on NAVSIM v2 and HUGSIM.
arXiv:2609.36851v1 Announce Type: new Abstract: End-to-end autonomous driving policies are commonly trained via imitation learning on logged demonstrations without observing the consequences of their...
arXiv:2610.32222v1 Announce Type: new Abstract: Reinforcement learning has gained attention for improving action-conditioned world models (ACWMs), as it enables reward-driven optimization of visual q...