DynaFLIP: Rethinking Robotics Perception via Tri-Modal-Dynamics Guided Representation

• URL: http://arxiv.org/abs/2605.30350v1
• Date: Thu, 28 May 2026 17:59:53 GMT
• Title: DynaFLIP: Rethinking Robotics Perception via Tri-Modal-Dynamics Guided Representation
• Abstract summary: We introduce DynaFLIP, a multimodal pre-training framework that pushes motion understanding upstream into perception.<n>We construct image-language-3D flow triplets from heterogeneous human and robot videos, and use these triplets as training-time supervision to shape an image-only encoder.<n>Our results suggest that robot generalization improves when visual representations are trained to encode not just what is present, but how the world changes under action.
• Score: 54.860960182207236
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Robot manipulation critically depends on perception that preserves the action-relevant aspects of a scene. Yet most robot learning pipelines are built upon visual encoders pre-trained for static recognition or vision-language alignment, leaving motion understanding to downstream policies. We introduce DynaFLIP, a dynamics-aware multimodal pre-training framework that pushes motion understanding upstream into perception. We construct image-language-3D flow triplets from heterogeneous human and robot videos, and use these triplets as training-time supervision to shape an image-only encoder. Our key idea is to encourage the three modalities to span a small simplex volume in the shared hyperspherical space -- a smaller simplex volume indicating stronger alignment. To avoid the geometric ambiguity and trivial collapse of naive volume minimization, we combine simplex-volume minimization with a cosine regularizer and a contrastive objective. Our analyses show that DynaFLIP focuses on control-relevant regions critical for manipulation. The resulting dynamics-aware representations serve as reusable visual backbones and consistently outperform baselines across diverse downstream policies, including VLAs. We validate this across diverse simulation and real-world setups, with gains reaching +22.5% under out-of-distribution scenarios. Our results suggest that robot generalization improves when visual representations are trained to encode not just what is present, but how the world changes under action.

Related papers

• LAWM-3D: Learning 3D-Aware Latent Actions from Human Videos for Generalizable Robot World Models [64.11117186468533]
  World models enable agents to perform forward rollout and planning without real-world interaction.<n>Most existing LAMs rely on single-view inputs and operate primarily in 2D pixel space.<n>We propose LAWM-3D, which introduces three tightly coupled key designs for learning 3D-aware latent actions.
  arXiv  Detail & Related papers  (2026-08-06T07:47:01Z)
• Contrastive Action-Image Pre-training for Visuomotor Control [91.5630572541332]
  We introduce CAIP, a vision encoder that treats human hand poses from large-scale egocentric video as a proxy for end-effector actions.<n>Leveraging 32,041 hours of egocentric human video and only 88 hours of robotic manipulation data, CAIP outperforms state-of-the-art vision encoders.
  arXiv  Detail & Related papers  (2026-06-15T20:00:20Z)
• Learning Action Manifold with Multi-view Latent Priors for Robotic Manipulation [67.16145181790522]
  This paper tackles spatial perception and manipulation challenges in Vision-Language-Action (VLA) models.<n>We propose a Geometry-Guided Gated Transformer (G3T) that aligns multi-view features under 3D geometric guidance.<n>To improve action learning efficiency, we introduce Action Manifold Learning (AML), which directly predicts actions on the valid action manifold.
  arXiv  Detail & Related papers  (2026-05-12T09:21:29Z)
• StereoPolicy: Improving Robotic Manipulation Policies via Stereo Perception [30.791524912385924]
  StereoPolicy is a visuomotor policy learning framework that directly leverages synchronized stereo image pairs to strengthen geometric reasoning.<n>Our results underscore stereo vision as a scalable and robust modality that bridges 2D pretrained representations with 3D geometric understanding for robotic manipulation.
  arXiv  Detail & Related papers  (2026-05-11T05:06:12Z)
• Cortical Policy: A Dual-Stream View Transformer for Robotic Manipulation [57.28703268044067]
  We propose Cortical Policy, a novel dual-stream view transformer for robotic manipulation.<n>Our framework offers a fresh perspective for robotic manipulation and holds potential for broader application in vision-based robot control.
  arXiv  Detail & Related papers  (2026-03-22T04:18:54Z)
• 3PoinTr: 3D Point Tracks for Robot Manipulation Pretraining from Casual Videos [8.359830715928242]
  3PoinTr is a method for pretraining robot policies from casual and unconstrained human videos.<n>3PoinTr uses a transformer architecture to predict 3D point tracks as an intermediate embodiment-agnostic representation.<n>We find that 3PoinTr produces more accurate and higher quality point tracks due to a lightweight yet expressive architecture.
  arXiv  Detail & Related papers  (2026-03-09T15:20:26Z)
• DynaRend: Learning 3D Dynamics via Masked Future Rendering for Robotic Manipulation [52.136378691610524]
  We present DynaRend, a representation learning framework that learns 3D-aware and dynamics-informed triplane features.<n>By pretraining on multi-view RGB-D video data, DynaRend jointly captures spatial geometry, future dynamics, and task semantics in a unified triplane representation.<n>We evaluate DynaRend on two challenging benchmarks, RLBench and Colosseum, demonstrating substantial improvements in policy success rate, generalization to environmental perturbations, and real-world applicability across diverse manipulation tasks.
  arXiv  Detail & Related papers  (2025-10-28T10:17:11Z)
• Zero-Shot Visual Generalization in Robot Manipulation [0.13280779791485384]
  Current approaches often sidestep the problem by relying on invariant representations such as point clouds and depth.<n>Disentangled representation learning has recently shown promise in enabling vision-based reinforcement learning policies to be robust to visual distribution shifts.<n>We demonstrate zero-shot adaptability to visual perturbations in both simulation and on real hardware.
  arXiv  Detail & Related papers  (2025-05-16T22:01:46Z)
• Human-oriented Representation Learning for Robotic Manipulation [64.59499047836637]
  Humans inherently possess generalizable visual representations that empower them to efficiently explore and interact with the environments in manipulation tasks. We formalize this idea through the lens of human-oriented multi-task fine-tuning on top of pre-trained visual encoders. Our Task Fusion Decoder consistently improves the representation of three state-of-the-art visual encoders for downstream manipulation policy-learning.
  arXiv  Detail & Related papers  (2023-10-04T17:59:38Z)
• 3D Neural Scene Representations for Visuomotor Control [78.79583457239836]
  We learn models for dynamic 3D scenes purely from 2D visual observations. A dynamics model, constructed over the learned representation space, enables visuomotor control for challenging manipulation tasks.
  arXiv  Detail & Related papers  (2021-07-08T17:49:37Z)

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.