mlsanityData to 7 Oct 2026

SCARP: 3D Shape Completion in ARbitrary Poses for Improved Grasping

arXiv 2301.07213 ↗First posted 17 Jan 2023cs.CV cs.ROcited by 8

Not on the map. Why?

Abstract (arXiv)

Recovering full 3D shapes from partial observations is a challenging task that has been extensively addressed in the computer vision community. Many deep learning methods tackle this problem by training 3D shape generation networks to learn a prior over the full 3D shapes. In this training regime, the methods expect the inputs to be in a fixed canonical form, without which they fail to learn a valid prior over the 3D shapes. We propose SCARP, a model that performs Shape Completion in ARbitrary Poses. Given a partial pointcloud of an object, SCARP learns a disentangled feature representation of pose and shape by relying on rotationally equivariant pose features and geometric shape features trained using a multi-tasking objective. Unlike existing methods that depend on an external canonicalization, SCARP performs canonicalization, pose estimation, and shape completion in a single network, improving the performance by 45% over the existing baselines. In this work, we use SCARP for improving grasp proposals on tabletop objects. By completing partial tabletop objects directly in their observed poses, SCARP enables a SOTA grasp proposal network improve their proposals by 71.2% on partial shapes. Project page: https://bipashasen.github.io/scarp

Key facts. Cited by 8 papers in the map (data through 7 Oct 2026). Builds on 39 of its 53 references that are in the map. First posted on arXiv 17 Jan 2023.

Builds on

39 of 53 references.

References · 39

  1. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  2. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,384
  3. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,880
  4. 2019DeepSDF: Learning Continuous Signed Distance Functions for Shape Representationcited by 2,682
  5. 2018Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point cloudscited by 735
  6. 2019Normalized Object Coordinate Space for Category-Level 6D Object Pose and Size Estimationcited by 490
  7. 2016Semantic Scene Completion from a Single Depth Imagecited by 468
  8. 2018PCN: Point Completion Networkcited by 382
  9. 2021GAN Inversion: A Surveycited by 366
  10. 20183D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Datacited by 348
  11. 2021PoinTr: Diverse Point Cloud Completion with Geometry-Aware Transformerscited by 284
  12. 2019Cormorant: Covariant Molecular Neural Networkscited by 262
27 more
  1. 2017Dex-Net 2.0: Deep Learning to Plan Robust Grasps with Synthetic Point Clouds and Analytic Grasp Metricscited by 259
  2. 20196-DOF GraspNet: Variational Grasp Generation for Object Manipulationcited by 237
  3. 2016Shape Completion using 3D-Encoder-Predictor CNNs and Shape Synthesiscited by 219
  4. 2020MetaSDF: Meta-learning Signed Distance Functionscited by 214
  5. 2021Contact-GraspNet: Efficient 6-DoF Grasp Generation in Cluttered Scenescited by 213
  6. 2022AutoSDF: Shape Priors for 3D Completion, Reconstruction and Generationcited by 208
  7. 2020GRNet: Gridding Residual Network for Dense Point Cloud Completioncited by 174
  8. 20193D Point Cloud Generative Adversarial Network Based on Tree Structured Graph Convolutionscited by 173
  9. 2020Cascaded Refinement Network for Point Cloud Completioncited by 112
  10. 2021Density-aware Chamfer Distance as a Comprehensive Metric for Point Cloud Completioncited by 111
  11. 2022ShapeFormer: Transformer-based Shape Completion via Sparse Representationcited by 106
  12. 2017High-Resolution Shape Completion Using Deep Neural Networks for Global Structure and Local Geometry Inferencecited by 87
  13. 2022SeedFormer: Patch Seeds based Point Cloud Completion with Upsample Transformercited by 83
  14. 2021Unsupervised 3D Shape Completion through GAN Inversioncited by 70
  15. 2020DSDNet: Deep Structured self-Driving Networkcited by 61
  16. 2020A Wigner-Eckart Theorem for Group Equivariant Convolution Kernelscited by 46
  17. 2019End-to-End CAD Model Retrieval and 9DoF Alignment in 3D Scanscited by 40
  18. 2020SoftPoolNet: Shape Descriptor for Point Cloud Completion and Classificationcited by 33
  19. 2020Weakly-supervised 3D Shape Completion in the Wildcited by 33
  20. 2022ConDor: Self-Supervised Canonicalization of 3D Pose for Partial Shapescited by 31
  21. 2020Learning to Orient Surfaces by Self-supervised Spherical CNNscited by 26
  22. 2020Same Object, Different Grasps: Data and Semantic Knowledge for Task-Oriented Graspingcited by 26
  23. 2020Cascaded Refinement Network for Point Cloud Completion with Self-supervisioncited by 20
  24. 2020Point Cloud Completion by Learning Shape Priorscited by 14
  25. 2021Multi-View Partial (MVP) Point Cloud Challenge 2021 on Completion and Registration: Methods and Resultscited by 9
  26. 2022Temporal Point Cloud Completion with Pose Disturbancecited by 5
  27. 2022Approaches and Challenges in Robotic Perception for Table-top Rearrangement and Planningcited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 2; 2024: 1; 2025: 3; 2026: 2.

8 papers.

  1. 2023HyP-NeRF: Learning Improved NeRF Priors using a HyperNetworkcited by 14
  2. 2024ESCAPE: Equivariant Shape Completion via Anchor Point Encodingcited by 4
  3. 20233D Surface Reconstruction in the Wild by Deforming Shape Priors from Synthetic Datacited by 2
  4. 2025Learning Generalizable Shape Completion with SIM(3) Equivariancecited by 2
  5. 2025Point-based Instance Completion with Scene Constraintscited by 1
  6. 2025A Multi-Level Similarity Approach for Single-View Object Grasping: Matching, Planning, and Fine-Tuningcited by 0
  7. 2026REVNET: Rotation-Equivariant Point Cloud Completion via Vector Neuron Anchor Transformercited by 0
  8. 2026Object Pose and Shape Estimation for Grasping: Does it Work?cited by 0