Dyn-3D: Unveiling and Resolving Ego-Motion Ambiguity in Vision-Language Models

• URL: http://arxiv.org/abs/2609.01059v1
• Date: Tue, 01 Sep 2026 10:54:27 GMT
• Title: Dyn-3D: Unveiling and Resolving Ego-Motion Ambiguity in Vision-Language Models
• Abstract summary: We introduce Dyn-3D, a benchmark using counterfactual 3D rendering to rigorously decouple visual changes from true kinematic properties.<n>By embedding metric physical ground truth into policy optimization, TempoVista explicitly grounds visual representations in 3D space.
• Score: 11.352994779962208
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: As Vision-Language Models (VLMs) tackle dynamic 3D spatial reasoning, ego-motion perception becomes essential to resolve monocular scale ambiguity. However, current models often overfit to smooth trajectory priors rather than genuinely understanding physical motion. Consequently, their spatial reasoning degrades severely under large displacements, a phenomenon we term Kinematic Collapse. This failure stems from spurious visual-motion correlations in natural videos and a lack of explicit physical supervision. To evaluate this, we introduce Dyn-3D, a benchmark using counterfactual 3D rendering to rigorously decouple visual changes from true kinematic properties. Furthermore, we propose the TempoVista framework, featuring the Kinematic-GSPO algorithm. By embedding metric physical ground truth into policy optimization, TempoVista explicitly grounds visual representations in 3D space. Experiments demonstrate that our approach significantly improves both motion estimation and robust spatial reasoning by utilizing camera dynamics as an effective geometric calibration signal.

Related papers

• Neural Voxel Dynamics: Learning Implicit 3D Physics via Volumetric Feature Advection [0.14323566945483496]
  We present a self-dimensional framework for learning implicit 3D physical dynamics directly from video-derived supervisory signals.<n>We address this by shifting the predictive bottleneck from 2D image space to a lifted' 3D Volumetric Latent Space.<n>Unlike state-of-the-art hybrid models that rely on explicit classical simulators for training and/or inference, our architecture tracks material states implicitly within high-supervised V-JEPA features.
  arXiv  Detail & Related papers  (2026-06-24T22:06:35Z)
• Kinematics-Driven Gaussian Shape Deformation for Blurry Monocular Dynamic Scenes [32.40228197424996]
  We present Kinematics-GS, a kinematics-aware framework that models blur as motion-aligned deformation.<n>To stabilize optimization, we decompose scenes into dynamic and static components using temporal deformation variance.<n>We also introduce a challenging real-world dataset of deformable and elastic objects exhibiting non-rigid motion.
  arXiv  Detail & Related papers  (2026-05-09T03:01:17Z)
• MASS: Motion-Aware Spatial-Temporal Grounding for Physics Reasoning and Comprehension in Vision-Language Models [45.450035386882824]
  Vision Language Models (VLMs) perform well on standard video tasks but struggle with physics-driven reasoning involving motion dynamics and spatial interactions.<n>We present an approach that addresses this gap by translating physical-world context cues into interpretable representations aligned with VLMs' perception, comprehension, and reasoning.
  arXiv  Detail & Related papers  (2025-11-23T09:43:44Z)
• PhysHMR: Learning Humanoid Control Policies from Vision for Physically Plausible Human Motion Reconstruction [52.44375492811009]
  We present PhysHMR, a unified framework that learns a visual-to-action policy for humanoid control in a physics-based simulator.<n>A key component of our approach is the pixel-as-ray strategy, which lifts 2D keypoints into 3D spatial rays and transforms them into global space.<n>PhysHMR produces high-fidelity, physically plausible motion across diverse scenarios, outperforming prior approaches in both visual accuracy and physical realism.
  arXiv  Detail & Related papers  (2025-10-02T21:01:11Z)
• Forecasting Future Videos from Novel Views via Disentangled 3D Scene Representation [54.60804602905519]
  We learn an entangled representation, aiming to model layered scene geometry, motion forecasting and novel view synthesis together. Our approach chooses to disentangle scene geometry from scene motion, via lifting the 2D scene to 3D point clouds. To model future 3D scene motion, we propose a disentangled two-stage approach that initially forecasts ego-motion and subsequently the residual motion of dynamic objects.
  arXiv  Detail & Related papers  (2024-07-31T08:54:50Z)
• Shape of Motion: 4D Reconstruction from a Single Video [42.42669078777769]
  We introduce a method for reconstructing generic dynamic scenes, featuring explicit, persistent 3D motion trajectories in the world coordinate frame.<n>First, we exploit the low-dimensional structure of 3D motion by representing scene motion with a compact set of SE(3) motion bases.<n>Second, we take advantage of off-the-shelf data-driven priors such as monocular depth maps and long-range 2D tracks, and devise a method to effectively consolidate these noisy supervisory signals.
  arXiv  Detail & Related papers  (2024-07-18T17:59:08Z)
• DO3D: Self-supervised Learning of Decomposed Object-aware 3D Motion and Depth from Monocular Videos [76.01906393673897]
  We propose a self-supervised method to jointly learn 3D motion and depth from monocular videos. Our system contains a depth estimation module to predict depth, and a new decomposed object-wise 3D motion (DO3D) estimation module to predict ego-motion and 3D object motion. Our model delivers superior performance in all evaluated settings.
  arXiv  Detail & Related papers  (2024-03-09T12:22:46Z)
• Decoupling Dynamic Monocular Videos for Dynamic View Synthesis [50.93409250217699]
  We tackle the challenge of dynamic view synthesis from dynamic monocular videos in an unsupervised fashion. Specifically, we decouple the motion of the dynamic objects into object motion and camera motion, respectively regularized by proposed unsupervised surface consistency and patch-based multi-view constraints.
  arXiv  Detail & Related papers  (2023-04-04T11:25:44Z)
• 3D-OES: Viewpoint-Invariant Object-Factorized Environment Simulators [24.181604511269096]
  We propose an action-conditioned dynamics model that predicts scene changes caused by object and agent interactions in a viewpoint-in 3D neural scene representation space. In this space, objects do not interfere with one another and their appearance persists over time and across viewpoints. We show our model generalizes well its predictions across varying number and appearances of interacting objects as well as across camera viewpoints.
  arXiv  Detail & Related papers  (2020-11-12T16:15:52Z)
• Kinematic 3D Object Detection in Monocular Video [123.7119180923524]
  We propose a novel method for monocular video-based 3D object detection which carefully leverages kinematic motion to improve precision of 3D localization. We achieve state-of-the-art performance on monocular 3D object detection and the Bird's Eye View tasks within the KITTI self-driving dataset.
  arXiv  Detail & Related papers  (2020-07-19T01:15:12Z)
• Motion Guided 3D Pose Estimation from Videos [81.14443206968444]
  We propose a new loss function, called motion loss, for the problem of monocular 3D Human pose estimation from 2D pose. In computing motion loss, a simple yet effective representation for keypoint motion, called pairwise motion encoding, is introduced. We design a new graph convolutional network architecture, U-shaped GCN (UGCN), which captures both short-term and long-term motion information.
  arXiv  Detail & Related papers  (2020-04-29T06:59:30Z)

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.