mlsanityData to 14 Oct 2026

Shape-Pose Disentanglement using SE(3)-equivariant Vector Neurons

arXiv 2204.01159 ↗First posted 3 Apr 2022cs.CVcited by 16

Not on the map. Why?

Abstract (arXiv)

We introduce an unsupervised technique for encoding point clouds into a canonical shape representation, by disentangling shape and pose. Our encoder is stable and consistent, meaning that the shape encoding is purely pose-invariant, while the extracted rotation and translation are able to semantically align different input shapes of the same class to a common canonical pose. Specifically, we design an auto-encoder based on Vector Neuron Networks, a rotation-equivariant neural network, whose layers we extend to provide translation-equivariance in addition to rotation-equivariance only. The resulting encoder produces pose-invariant shape encoding by construction, enabling our approach to focus on learning a consistent canonical pose for a class of objects. Quantitative and qualitative experiments validate the superior stability and consistency of our approach.

Key facts. Cited by 16 papers in the map (data through 14 Oct 2026). Builds on 28 of its 31 references that are in the map. First posted on arXiv 3 Apr 2022.

Builds on

28 of 31 references.

References · 28

  1. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,248
  2. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,401
  3. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,897
  4. 2019DeepSDF: Learning Continuous Signed Distance Functions for Shape Representationcited by 2,689
  5. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,223
  6. 2018Occupancy Networks: Learning 3D Reconstruction in Function Spacecited by 2,041
  7. 2018PointRCNN: 3D Object Proposal Generation and Detection from Point Cloudcited by 941
  8. 2018Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point cloudscited by 746
  9. 20163D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstructioncited by 588
  10. 2018AtlasNet: A Papier-Mach'e Approach to Learning 3D Surface Generationcited by 539
  11. 2018Spherical CNNscited by 456
  12. 20203DSSD: Point-based 3D Single Stage Object Detectorcited by 440
16 more
  1. 2022Rethinking Network Design and Local Geometry in Point Cloud: A Simple Residual MLP Frameworkcited by 379
  2. 20183D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Datacited by 354
  3. 2019DISN: Deep Implicit Surface Network for High-quality Single-view 3D Reconstructioncited by 333
  4. 2016Steerable CNNscited by 305
  5. 2021Vector Neurons: A General Framework for SO(3)-Equivariant Networkscited by 305
  6. 2017Learning SO(3) Equivariant Representations with Spherical CNNscited by 248
  7. 2020MVTN: Multi-View Transformation Network for 3D Shape Recognitioncited by 140
  8. 2019TANet: Robust 3D Object Detection from Point Clouds with Triple Attentioncited by 108
  9. 2019Learning elementary structures for 3D shape generation and matchingcited by 98
  10. 2019C3DPO: Canonical 3D Pose Networks for Non-Rigid Structure From Motioncited by 61
  11. 2020CaSPR: Learning Canonical Spatiotemporal Point Cloud Representationscited by 51
  12. 2021Leveraging SE(3) Equivariance for Self-Supervised Category-Level Object Pose Estimationcited by 45
  13. 2020Weakly-supervised 3D Shape Completion in the Wildcited by 33
  14. 2022ConDor: Self-Supervised Canonicalization of 3D Pose for Partial Shapescited by 31
  15. 2020Canonical Capsules: Self-Supervised Capsules in Canonical Posecited by 30
  16. 2020Learning to Orient Surfaces by Self-supervised Spherical CNNscited by 26

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 1; 2023: 8; 2024: 3; 2025: 2; 2026: 2.

16 papers.

Sort:

  1. 2023EquivAct: SIM(3)-Equivariant Visuomotor Policies beyond Rigid Object Manipulationcited by 37
  2. 2023Leveraging SE(3) Equivariance for Learning 3D Geometric Shape Assemblycited by 28
  3. 2023Banana: Banach Fixed-Point Network for Pointcloud Segmentation with Inter-Part Equivariancecited by 19
  4. 2023EFEM: Equivariant Neural Field Expectation Maximization for 3D Object Segmentation Without Scene Supervisioncited by 18
  5. 2023Generalizing Neural Human Fitting to Unseen Poses With Articulated SE(3) Equivariancecited by 14
  6. 2023ShapeMatcher: Self-Supervised Joint Shape Canonicalization, Segmentation, Retrieval and Deformationcited by 8
  7. 2024Approximately Piecewise E(3) Equivariant Point Networkscited by 4
  8. 2025CrossJEPA: Cross-Modal Joint-Embedding Predictive Architecture for Efficient 3D Representation Learning from 2D Imagescited by 4
  9. 2022Category-Level Global Camera Pose Estimation with Multi-Hypothesis Point Cloud Correspondencescited by 3
  10. 2023Stable and Consistent Prediction of 3D Characteristic Orientation via Invariant Residual Learningcited by 3
  11. 2024Learning SO(3)-Invariant Semantic Correspondence via Local Shape Transformcited by 3
  12. 20233D Surface Reconstruction in the Wild by Deforming Shape Priors from Synthetic Datacited by 2
4 more
  1. 2025Learning Generalizable Shape Completion with SIM(3) Equivariancecited by 2
  2. 2024Symmetry-Robust 3D Orientation Estimationcited by 1
  3. 2026Coalgebras for categorical deep learning: Representability and universal approximationcited by 0
  4. 2026SCAPO: Self-Supervised Category-Level Articulated Pose Estimation from a Single 3D Observationcited by 0