Learning Geometrically-Grounded 3D Visual Representations for View-Generalizable Robotic Manipulation

• URL: http://arxiv.org/abs/2601.22988v1
• Date: Fri, 30 Jan 2026 13:53:53 GMT
• Title: Learning Geometrically-Grounded 3D Visual Representations for View-Generalizable Robotic Manipulation
• Abstract summary: Real-world robotic manipulation demands robust spatial scene understanding and strong generalization across diverse camera viewpoints.<n>Recent advances in 3D-aware visual representations have shown promise, but they still suffer from several key limitations.<n>We present MethodName, a unified representation-policy learning framework for view-generalizable robotic manipulation.
• Score: 19.812030036563335
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Real-world robotic manipulation demands visuomotor policies capable of robust spatial scene understanding and strong generalization across diverse camera viewpoints. While recent advances in 3D-aware visual representations have shown promise, they still suffer from several key limitations, including reliance on multi-view observations during inference which is impractical in single-view restricted scenarios, incomplete scene modeling that fails to capture holistic and fine-grained geometric structures essential for precise manipulation, and lack of effective policy training strategies to retain and exploit the acquired 3D knowledge. To address these challenges, we present MethodName, a unified representation-policy learning framework for view-generalizable robotic manipulation. MethodName introduces a single-view 3D pretraining paradigm that leverages point cloud reconstruction and feed-forward gaussian splatting under multi-view supervision to learn holistic geometric representations. During policy learning, MethodName performs multi-step distillation to preserve the pretrained geometric understanding and effectively transfer it to manipulation skills. We conduct experiments on 12 RLBench tasks, where our approach outperforms the previous state-of-the-art method by 12.7% in average success rate. Further evaluation on six representative tasks demonstrates strong zero-shot view generalization, with success rate drops of only 22.0% and 29.7% under moderate and large viewpoint shifts respectively, whereas the state-of-the-art method suffers larger decreases of 41.6% and 51.5%.

Related papers

• 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)
• Generalizable Geometric Prior and Recurrent Spiking Feature Learning for Humanoid Robot Manipulation [90.90219129619344]
  This paper presents a novel R-prior-S, Recurrent Geometric-priormodal Policy with Spiking features.<n>To ground high-level reasoning in physical reality, we leverage lightweight 2D geometric inductive biases.<n>For the data efficiency issue in robotic action generation, we introduce a Recursive Adaptive Spiking Network.
  arXiv  Detail & Related papers  (2026-01-13T23:36:30Z)
• Opening the Sim-to-Real Door for Humanoid Pixel-to-Action Policy Transfer [59.02729900344616]
  GPU-accelerated, photorealistic simulation has opened a scalable data-generation path for robot learning.<n>We develop a teacher-student-bootstrap learning framework for vision-based humanoid loco-manipulation.<n>This represents the first humanoid sim-to-real policy capable of diverse articulated loco-manipulation using pure RGB perception.
  arXiv  Detail & Related papers  (2025-11-30T20:07:13Z)
• 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)
• ManiVID-3D: Generalizable View-Invariant Reinforcement Learning for Robotic Manipulation via Disentangled 3D Representations [19.02933938928656]
  ManiVID-3D is a novel 3D visual reinforcement learning architecture for robotic manipulation.<n>It learns view-invariant representations through self-supervised disentangled feature learning.<n>It achieves a 44.7% higher success rate than state-of-the-art methods under viewpoint variations.
  arXiv  Detail & Related papers  (2025-09-14T06:31:04Z)
• CL3R: 3D Reconstruction and Contrastive Learning for Enhanced Robotic Manipulation Representations [19.71090711790973]
  We propose a novel 3D pre-training framework designed to enhance robotic manipulation policies.<n>Our method integrates both spatial awareness and semantic understanding by employing a point cloud Masked Autoencoder.<n>We mitigate camera view ambiguity and improve generalization, enabling robust perception from novel viewpoints at test time.
  arXiv  Detail & Related papers  (2025-07-11T02:16:32Z)
• Learning Precise Affordances from Egocentric Videos for Robotic Manipulation [25.929092988536087]
  Affordance, defined as the potential actions that an object offers, is crucial for embodied AI agents.<n>We propose a complete affordance learning system that takes in egocentric videos and outputs precise affordance annotations without human labeling.<n>We also introduce a framework that facilitates affordance-oriented robotic manipulation such as tool grasping and robot-to-human tool handover.
  arXiv  Detail & Related papers  (2024-08-19T16:11:47Z)
• 3D Diffusion Policy: Generalizable Visuomotor Policy Learning via Simple 3D Representations [19.41216557646392]
  3D Diffusion Policy (DP3) is a novel visual imitation learning approach. In experiments, DP3 handles most tasks with just 10 demonstrations and surpasses baselines with a 24.2% relative improvement. In real robot experiments, DP3 rarely violates safety requirements, in contrast to baseline methods which frequently do.
  arXiv  Detail & Related papers  (2024-03-06T18:58:49Z)
• Self-Supervised Learning of Multi-Object Keypoints for Robotic Manipulation [8.939008609565368]
  In this paper, we demonstrate the efficacy of learning image keypoints via the Dense Correspondence pretext task for downstream policy learning. We evaluate our approach on diverse robot manipulation tasks, compare it to other visual representation learning approaches, and demonstrate its flexibility and effectiveness for sample-efficient policy learning.
  arXiv  Detail & Related papers  (2022-05-17T13:15:07Z)
• On Triangulation as a Form of Self-Supervision for 3D Human Pose Estimation [57.766049538913926]
  Supervised approaches to 3D pose estimation from single images are remarkably effective when labeled data is abundant. Much of the recent attention has shifted towards semi and (or) weakly supervised learning. We propose to impose multi-view geometrical constraints by means of a differentiable triangulation and to use it as form of self-supervision during training when no labels are available.
  arXiv  Detail & Related papers  (2022-03-29T19:11:54Z)
• A Framework for Efficient Robotic Manipulation [79.10407063260473]
  We show that a single robotic arm can learn sparse-reward manipulation policies from pixels. We show that, given only 10 demonstrations, a single robotic arm can learn sparse-reward manipulation policies from pixels.
  arXiv  Detail & Related papers  (2020-12-14T22:18:39Z)
• Weakly-Supervised 3D Human Pose Learning via Multi-view Images in the Wild [101.70320427145388]
  We propose a weakly-supervised approach that does not require 3D annotations and learns to estimate 3D poses from unlabeled multi-view data. We evaluate our proposed approach on two large scale datasets.
  arXiv  Detail & Related papers  (2020-03-17T08:47:16Z)

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.