ShapeGen: Robotic Data Generation for Category-Level Manipulation

• URL: http://arxiv.org/abs/2604.15569v1
• Date: Thu, 16 Apr 2026 22:55:39 GMT
• Title: ShapeGen: Robotic Data Generation for Category-Level Manipulation
• Abstract summary: ShapeGen is a method to generate 3D manipulation data in a simulator-free and 3D manner.<n>In the first stage, we train spatial warpings between shapes mapping points to points that correspond functionally, and aggregate 3D models along with the warpings into a plug-and-play Shape Library.<n>In the second stage, we design a pipeline that, leveraging established Libraries, requires only minimal human annotation to generate physically plausible and functionally correct demonstrations.
• Score: 57.10250198611112
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Manipulation policies deployed in uncontrolled real-world scenarios are faced with great in-category geometric diversity of everyday objects. In order to function robustly under such variations, policies need to work in a category-level manner, i.e. knowing how to interact with any object in a certain category, instead of only a specific one seen during training. This in-category generalizability is usually nurtured with shape-diversified training data; however, manually collecting such a corpus of data is infeasible due to the requirement of intense human labor and large collections of divergent objects at hand. In this paper, we propose ShapeGen, a data generation method that aims at generating shape-variated manipulation data in a simulator-free and 3D manner. ShapeGen decomposes the process into two stages: Shape Library curation and Function-Aware Generation. In the first stage, we train spatial warpings between shapes mapping points to points that correspond functionally, and aggregate 3D models along with the warpings into a plug-and-play Shape Library. In the second stage, we design a pipeline that, leveraging established Libraries, requires only minimal human annotation to generate physically plausible and functionally correct novel demonstrations. Experiments in the real world demonstrate the effectiveness of ShapeGen to boost policies' in-category shape generalizability. Project page: https://wangyr22.github.io/ShapeGen/.

Related papers

• 3D Shape Completion on Unseen Categories:A Weakly-supervised Approach [61.76304400106871]
  We introduce a novel weakly-supervised framework to reconstruct the complete shapes from unseen categories. We first propose an end-to-end prior-assisted shape learning network that leverages data from the seen categories to infer a coarse shape. In addition, we propose a self-supervised shape refinement model to further refine the coarse shape.
  arXiv  Detail & Related papers  (2024-01-19T09:41:09Z)
• Generalizable Local Feature Pre-training for Deformable Shape Analysis [36.44119664239748]
  Transfer learning is fundamental for addressing problems in settings with little training data. We analyze the link between feature locality and transferability in tasks involving deformable 3D objects. We propose a differentiable method for optimizing the receptive field within 3D transfer learning.
  arXiv  Detail & Related papers  (2023-03-27T11:13:46Z)
• Generative Category-Level Shape and Pose Estimation with Semantic Primitives [27.692997522812615]
  We propose a novel framework for category-level object shape and pose estimation from a single RGB-D image. To handle the intra-category variation, we adopt a semantic primitive representation that encodes diverse shapes into a unified latent space. We show that the proposed method achieves SOTA pose estimation performance and better generalization in the real-world dataset.
  arXiv  Detail & Related papers  (2022-10-03T17:51:54Z)
• Single-view 3D Mesh Reconstruction for Seen and Unseen Categories [69.29406107513621]
  Single-view 3D Mesh Reconstruction is a fundamental computer vision task that aims at recovering 3D shapes from single-view RGB images. This paper tackles Single-view 3D Mesh Reconstruction, to study the model generalization on unseen categories. We propose an end-to-end two-stage network, GenMesh, to break the category boundaries in reconstruction.
  arXiv  Detail & Related papers  (2022-08-04T14:13:35Z)
• PatchRD: Detail-Preserving Shape Completion by Learning Patch Retrieval and Deformation [59.70430570779819]
  We introduce a data-driven shape completion approach that focuses on completing geometric details of missing regions of 3D shapes. Our key insight is to copy and deform patches from the partial input to complete missing regions. We leverage repeating patterns by retrieving patches from the partial input, and learn global structural priors by using a neural network to guide the retrieval and deformation steps.
  arXiv  Detail & Related papers  (2022-07-24T18:59:09Z)
• GenSDF: Two-Stage Learning of Generalizable Signed Distance Functions [37.07344649854001]
  We investigate the generalization capabilities of neural signed distance functions (SDFs) for learning 3D object representations for unseen and unlabeled point clouds. Existing methods can fit SDFs to a handful of object classes and boast fine detail or fast inference speeds, but do not generalize well to unseen shapes. We introduce a two-stage semi-supervised meta-learning approach that transfers shape priors from labeled to unlabeled data to reconstruct unseen object categories.
  arXiv  Detail & Related papers  (2022-06-06T17:58:29Z)
• GIFS: Neural Implicit Function for General Shape Representation [23.91110763447458]
  General Implicit Function for 3D Shape (GIFS) is a novel method to represent general shapes. Instead of dividing 3D space into predefined inside-outside regions, GIFS encodes whether two points are separated by any surface. Experiments on ShapeNet show that GIFS outperforms previous state-of-the-art methods in terms of reconstruction quality, rendering efficiency, and visual fidelity.
  arXiv  Detail & Related papers  (2022-04-14T17:29:20Z)
• NeuroMorph: Unsupervised Shape Interpolation and Correspondence in One Go [109.88509362837475]
  We present NeuroMorph, a new neural network architecture that takes as input two 3D shapes. NeuroMorph produces smooth and point-to-point correspondences between them. It works well for a large variety of input shapes, including non-isometric pairs from different object categories.
  arXiv  Detail & Related papers  (2021-06-17T12:25:44Z)
• Unsupervised 3D Shape Completion through GAN Inversion [116.27680045885849]
  We present ShapeInversion, which introduces Generative Adrial Network (GAN) inversion to shape completion for the first time. ShapeInversion uses a GAN pre-trained on complete shapes by searching for a latent code that gives a complete shape that best fits the given partial input. On the ShapeNet benchmark, the proposed ShapeInversion outperforms the SOTA unsupervised method, and is comparable with supervised methods that are learned using paired data.
  arXiv  Detail & Related papers  (2021-04-27T17:53:46Z)
• Deep Geometric Functional Maps: Robust Feature Learning for Shape Correspondence [31.840880075039944]
  We present a novel learning-based approach for computing correspondences between non-rigid 3D shapes. Key to our method is a feature-extraction network that learns directly from raw shape geometry.
  arXiv  Detail & Related papers  (2020-03-31T15:20:17Z)

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.