mlsanityData to 14 Oct 2026

Image to Icosahedral Projection for SO(3) Object Reasoning from Single-View Images

arXiv 2207.08925 ↗First posted 18 Jul 2022cs.CV cs.LGcited by 4

Not on the map. Why?

Abstract (arXiv)

Reasoning about 3D objects based on 2D images is challenging due to variations in appearance caused by viewing the object from different orientations. Tasks such as object classification are invariant to 3D rotations and other such as pose estimation are equivariant. However, imposing equivariance as a model constraint is typically not possible with 2D image input because we do not have an a priori model of how the image changes under out-of-plane object rotations. The only SO(3)-equivariant models that currently exist require point cloud or voxel input rather than 2D images. In this paper, we propose a novel architecture based on icosahedral group convolutions that reasons in SO(3) by learning a projection of the input image onto an icosahedron. The resulting model is approximately equivariant to rotation in SO(3). We apply this model to object pose estimation and shape classification tasks and find that it outperforms reasonable baselines. Project website: urlhttps://dmklee.github.io/image2icosahedral

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Builds on 34 of its 44 references that are in the map. First posted on arXiv 18 Jul 2022.

Builds on

34 of 44 references.

References · 34

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  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. 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,214
  5. 2019KPConv: Flexible and Deformable Convolution for Point Cloudscited by 1,205
  6. 2016Group Equivariant Convolutional Networkscited by 1,165
  7. 2017PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenescited by 775
  8. 2018Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point cloudscited by 746
  9. 2021E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentialscited by 708
  10. 2020SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networkscited by 607
  11. 2015Multi-view Convolutional Neural Networks for 3D Shape Recognitioncited by 565
  12. 2019Revisiting Point Cloud Classification: A New Benchmark Dataset and Classification Model on Real-World Datacited by 532
22 more
  1. 2019Normalized Object Coordinate Space for Category-Level 6D Object Pose and Size Estimationcited by 495
  2. 2018Spherical CNNscited by 456
  3. 2019General E(2)-Equivariant Steerable CNNscited by 446
  4. 2016Dynamic Filter Networkscited by 319
  5. 20203D-FUTURE: 3D Furniture shape with TextUREcited by 311
  6. 2021Vector Neurons: A General Framework for SO(3)-Equivariant Networkscited by 305
  7. 2019Gauge Equivariant Convolutional Networks and the Icosahedral CNNcited by 267
  8. 2017Learning SO(3) Equivariant Representations with Spherical CNNscited by 248
  9. 2018Deep Object Pose Estimation for Semantic Robotic Grasping of Household Objectscited by 241
  10. 2019PVN3D: A Deep Point-wise 3D Keypoints Voting Network for 6DoF Pose Estimationcited by 204
  11. 2019DPOD: 6D Pose Object Detector and Refinercited by 189
  12. 2019kPAM: KeyPoint Affordances for Category-Level Robotic Manipulationcited by 153
  13. 2021Provably Strict Generalisation Benefit for Equivariant Modelscited by 91
  14. 2021Equivariant Point Network for 3D Point Cloud Analysiscited by 81
  15. 2021Deep Regression on Manifolds: A 3D Rotation Case Studycited by 67
  16. 2018Explorations in Homeomorphic Variational Auto-Encodingcited by 65
  17. 2022Learning Symmetric Embeddings for Equivariant World Modelscited by 54
  18. 2020EfficientPose: An efficient, accurate and scalable end-to-end 6D multi object pose estimation approachcited by 46
  19. 2019Equivariant Multi-View Networkscited by 45
  20. 2019Spherical Regression: Learning Viewpoints, Surface Normals and 3D Rotations on n-Spherescited by 37
  21. 2016GIFT: A Real-time and Scalable 3D Shape Search Enginecited by 31
  22. 2016Orientation-boosted Voxel Nets for 3D Object Recognitioncited by 17

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 2; 2024: 1; 2025: 0; 2026: 1.

4 papers.

  1. 2023Equivariant Single View Pose Prediction Via Induced and Restricted Representationscited by 6
  2. 2023MELON: NeRF with Unposed Images in SO(3)cited by 4
  3. 20243D Equivariant Pose Regression via Direct Wigner-D Harmonics Predictioncited by 4
  4. 2026Equivariant Visual-Tactile Diffusion Policy for Contact-Rich Manipulationcited by 0