3D HAMSTER: Bridging Planning and Control in Hierarchical Vision Language Action Models through 3D Trajectory Guidance

• URL: http://arxiv.org/abs/2606.31329v2
• Date: Wed, 01 Jul 2026 13:05:08 GMT
• Title: 3D HAMSTER: Bridging Planning and Control in Hierarchical Vision Language Action Models through 3D Trajectory Guidance
• Abstract summary: 3D HAMSTER is a hierarchical framework that closes the gap between state-of-the-art low-level policies and metrically reliable 3D trajectories.<n>Across 3D trajectory prediction, simulation, and real-world manipulation, 3D HAMSTER consistently outperforms proprietary VLMs and 2D-guided baselines.
• Score: 46.721682793921815
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Hierarchical Vision-Language-Action (VLA) models decouple high-level planning from low-level control to improve generalization in robot manipulation. Recent work in this paradigm uses 2D end-effector trajectories predicted by a Vision-Language Model (VLM) as explicit guidance for a downstream policy. However, state-of-the-art low-level policies operate in 3D metric space on point clouds, and feeding them 2D guidance that lacks depth forces each waypoint to be assigned the depth of whatever scene surface lies beneath it, producing geometrically distorted trajectories. We propose 3D HAMSTER, a hierarchical framework that closes this gap by having the planner directly output metrically reliable 3D trajectories. We augment a VLM with a dedicated depth encoder and a dense depth reconstruction objective to predict 3D waypoint sequences, which are directly integrated into a pointcloudbased low-level policy. Across 3D trajectory prediction, simulation, and real-world manipulation, 3D HAMSTER consistently outperforms proprietary VLMs and 2D-guided baselines, with the largest gains under appearance-altering shifts and unseen language, spatial, and visual conditions. The project page is available at https://davian-robotics.github.io/3D_HAMSTER/.

Related papers

• 3DWay: Generalizing Robot Manipulation via 3D Consistent Waypoints [92.67359514804048]
  Intermediate representations are key to bridging the modality gap between generalizable manipulation policies and large-scale pretrained vision-language models.<n>We propose predicting 3D consistent waypoints (3DWay) from multi-view images.<n>By reformulating 3D waypoints prediction as generating multi-view consistent 2D waypoints followed by geometric triangulation, we enable explicit 3D motion specification.
  arXiv  Detail & Related papers  (2026-09-08T04:18:46Z)
• Lift3D-VLA: Lifting VLA Models to 3D Geometry and Dynamics-Aware Manipulation [76.16065370700488]
  We introduce Lift3D-VLA, a unified VLA framework that equips models with explicit 3D point cloud reasoning and enables temporally coherent action generation.<n>We show that Lift3D-VLA achieves 10.8% and 11.1% higher mean success rates on MetaWorld and RLBench than the best-performing prior VLA methods.
  arXiv  Detail & Related papers  (2026-07-07T17:59:47Z)
• PointACT: Vision-Language-Action Models with Multi-Scale Point-Action Interaction [57.63073414949329]
  We propose PointACT, a dual-system 3D-aware VLA policy that integrates hierarchical 3D point cloud representations directly into the action decoding process.<n>PointACT employs a multi-scale point-action interaction mechanism with efficient bottleneck window self-attention, enabling evolving action tokens to densely attend to both local geometric detail and global scene structure.
  arXiv  Detail & Related papers  (2026-05-20T17:10:31Z)
• ST-VLA: Enabling 4D-Aware Spatiotemporal Understanding for General Robot Manipulation [33.92613503473177]
  We propose a unified 3D-4D representation using a unified 3D-4D representation to bridge perception and action.<n>We introduce STHuman, a large-scale human manipulation dataset with 14 and 300k episodes, annotated 2D, 3D, and 4D contexts.<n>Experiments on RLBench and real-world manipulation tasks show that method significantly outperforms state-of-theart baselines.
  arXiv  Detail & Related papers  (2026-03-14T06:36:48Z)
• Action-Geometry Prediction with 3D Geometric Prior for Bimanual Manipulation [53.09168514034483]
  Bimanual manipulation requires policies that can reason about 3D geometry, anticipate how it evolves under action, and generate smooth, coordinated motions.<n>We propose a framework that builds bimanual manipulation directly on a pre-trained 3D geometric foundation model.<n>Our policy fuses geometry-aware latents, 2D semantic features, and proprioception into a unified state representation, and uses diffusion model to jointly predict a future action chunk and a future 3D latent that decodes into a dense pointmap.
  arXiv  Detail & Related papers  (2026-02-27T08:54:20Z)
• D3D-VLP: Dynamic 3D Vision-Language-Planning Model for Embodied Grounding and Navigation [66.7166217399105]
  Embodied agents face a critical dilemma that end-to-end models lack interpretability and explicit 3D reasoning.<n>Our model introduces two key innovations: 1) A Dynamic 3D Chain-of-Thought (3D CoT) that unifies planning, grounding, navigation, and question answering within a single 3D-VLM and CoT pipeline; 2) A Synergistic Learning from Fragmented Supervision (SLFS) strategy, which uses a masked autoregressive loss to learn from massive and partially-annotated hybrid data.
  arXiv  Detail & Related papers  (2025-12-14T09:53:15Z)
• GaussRender: Learning 3D Occupancy with Gaussian Rendering [86.89653628311565]
  GaussRender is a module that improves 3D occupancy learning by enforcing projective consistency.<n>Our method penalizes 3D configurations that produce inconsistent 2D projections, thereby enforcing a more coherent 3D structure.
  arXiv  Detail & Related papers  (2025-02-07T16:07:51Z)
• Lift3D Foundation Policy: Lifting 2D Large-Scale Pretrained Models for Robust 3D Robotic Manipulation [30.744137117668643]
  Lift3D is a framework that enhances 2D foundation models with implicit and explicit 3D robotic representations to construct a robust 3D manipulation policy.<n>In experiments, Lift3D consistently outperforms previous state-of-the-art methods across several simulation benchmarks and real-world scenarios.
  arXiv  Detail & Related papers  (2024-11-27T18:59:52Z)
• Act3D: 3D Feature Field Transformers for Multi-Task Robotic Manipulation [18.964403296437027]
  Act3D represents the robot's workspace using a 3D feature field with adaptive resolutions dependent on the task at hand. It samples 3D point grids in a coarse to fine manner, featurizes them using relative-position attention, and selects where to focus the next round of point sampling.
  arXiv  Detail & Related papers  (2023-06-30T17:34:06Z)
• Cylinder3D: An Effective 3D Framework for Driving-scene LiDAR Semantic Segmentation [87.54570024320354]
  State-of-the-art methods for large-scale driving-scene LiDAR semantic segmentation often project and process the point clouds in the 2D space. A straightforward solution to tackle the issue of 3D-to-2D projection is to keep the 3D representation and process the points in the 3D space. We develop a 3D cylinder partition and a 3D cylinder convolution based framework, termed as Cylinder3D, which exploits the 3D topology relations and structures of driving-scene point clouds.
  arXiv  Detail & Related papers  (2020-08-04T13:56:19Z)

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.