Future Dynamic 3D Reconstruction: A 3D World Model with Disentangled Ego-Motion

• URL: http://arxiv.org/abs/2606.18250v1
• Date: Tue, 16 Jun 2026 17:59:46 GMT
• Title: Future Dynamic 3D Reconstruction: A 3D World Model with Disentangled Ego-Motion
• Abstract summary: We propose FR3D, a world model that predicts a persistent 3D latent representation for future dynamic 3D reconstruction.<n>Experiments demonstrate FR3D's strong performance for future dynamic 3D reconstruction from monocular observations across multiple datasets, even 2 seconds into the future.
• Score: 69.61280478695205
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Forecasting the evolution of dynamic environments is crucial for autonomous agents. While generative world models have recently achieved high photorealism in 2D video synthesis by mixing ego-motion and environmental dynamics within the image plane, they exhibit physical inconsistencies, such as morphing or vanishing objects, especially over long time horizons. In this paper, we propose FR3D, a world model that predicts a persistent 3D latent representation for future dynamic 3D reconstruction. Unlike prior works that treat the world as a sequence of image-based features, FR3D explicitly decouples the 3D evolution of the scene from the agent's trajectory, treating the inferred ego-motion as a latent proxy for action. This disentanglement resolves the ambiguities between self-motion and world-motion, ensuring geometric consistency into the future. Furthermore, we introduce a teacher-student distillation strategy that leverages the spatial "common sense" of off-the-shelf foundation models, leading to robust zero-shot generalization. Extensive experiments demonstrate FR3D's strong performance for future dynamic 3D reconstruction from monocular observations across multiple datasets, even 2 seconds into the future. Project page: https://fr3d-wm.github.io.

Related papers

• 3D-Belief: Embodied Belief Inference via Generative 3D World Modeling [37.75852887428672]
  We present 3D-Belief, a generative 3D world model that infers explicit, actionable 3D beliefs from partial observations and updates them online over time.<n>Unlike prior visual prediction models, 3D-Belief represents uncertainty directly in 3D, enabling embodied agents to imagine plausible scene completions and reason over partially observed environments.
  arXiv  Detail & Related papers  (2026-05-12T00:42:27Z)
• Beyond Pixel Histories: World Models with Persistent 3D State [50.4601060508243]
  PERSIST is a new paradigm of world model which simulates the evolution of a latent 3D scene.<n>We show substantial improvements in spatial memory, 3D consistency, and long-horizon stability over existing methods.
  arXiv  Detail & Related papers  (2026-03-03T19:58:31Z)
• StemVLA:An Open-Source Vision-Language-Action Model with Future 3D Spatial Geometry Knowledge and 4D Historical Representation [6.0744834626758495]
  StemVLA is a novel framework that explicitly incorporates both future-oriented 3D spatial knowledge and historical 4D representations into action prediction.<n>We show that StemVLA significantly improves long-horizon task success and state-of-the-art performance on the CALVIN ABC-D benchmark [46], achieving an average sequence length of XXX.
  arXiv  Detail & Related papers  (2026-02-27T06:43:37Z)
• RAYNOVA: Scale-Temporal Autoregressive World Modeling in Ray Space [51.441415833480505]
  RAYNOVA is a multiview world model for driving scenarios that employs a dual-causal autoregressive framework.<n>It constructs an isotropic-temporal representation across views, frames, and scales based on relative Plcker-ray positional encoding.
  arXiv  Detail & Related papers  (2026-02-24T08:41:40Z)
• Terra: Explorable Native 3D World Model with Point Latents [74.90179419859415]
  We present Terra, a native 3D world model that represents and generates explorable environments in an intrinsic 3D latent space.<n>Specifically, we propose a novel point-to-Gaussian variational autoencoder (P2G-VAE) that encodes 3D inputs into a latent point representation.<n>We then introduce a sparse point flow matching network (SPFlow) for generating the latent point representation, which simultaneously denoises the positions and features of the point latents.
  arXiv  Detail & Related papers  (2025-10-16T17:59:56Z)
• From 2D to 3D Cognition: A Brief Survey of General World Models [16.121071388463694]
  3D-aware generative world models have demonstrated the ability to synthesize geometrically consistent, interactive 3D environments.<n>Despite rapid progress, the field lacks systematic analysis to categorize emerging techniques and clarify their roles in advancing 3D cognitive world models.<n>This survey addresses this need by introducing a conceptual framework, providing a structured and forward-looking review of world models transitioning from 2D perception to 3D cognition.
  arXiv  Detail & Related papers  (2025-06-25T05:05:09Z)
• Layered Motion Fusion: Lifting Motion Segmentation to 3D in Egocentric Videos [71.24593306228145]
  We propose to improve dynamic segmentation in 3D by fusing motion segmentation predictions from a 2D-based model into layered radiance fields.<n>We address this issue through test-time refinement, which helps the model to focus on specific frames, thereby reducing the data complexity.<n>This demonstrates that 3D techniques can enhance 2D analysis even for dynamic phenomena in a challenging and realistic setting.
  arXiv  Detail & Related papers  (2025-06-05T19:46:48Z)
• Learning 3D Persistent Embodied World Models [84.40585374179037]
  We introduce a new persistent embodied world model with an explicit memory of previously generated content.<n>During generation time, our video diffusion model predicts RGB-D video of the future observations of the agent.<n>This generation is then aggregated into a persistent 3D map of the environment.
  arXiv  Detail & Related papers  (2025-05-05T17:59:17Z)
• 3D-VLA: A 3D Vision-Language-Action Generative World Model [68.0388311799959]
  Recent vision-language-action (VLA) models rely on 2D inputs, lacking integration with the broader realm of the 3D physical world. We propose 3D-VLA by introducing a new family of embodied foundation models that seamlessly link 3D perception, reasoning, and action. Our experiments on held-in datasets demonstrate that 3D-VLA significantly improves the reasoning, multimodal generation, and planning capabilities in embodied environments.
  arXiv  Detail & Related papers  (2024-03-14T17:58:41Z)

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.