mlsanityData to 14 Oct 2026

3D Equivariant Visuomotor Policy Learning via Spherical Projection

arXiv 2505.16969 ↗First posted 22 May 2025cs.ROcited by 5

On the map: Diffusion models → Diffusion and flow policies for robotics

Abstract (arXiv)

Equivariant models have recently been shown to improve the data efficiency of diffusion policy by a significant margin. However, prior work that explored this direction focused primarily on point cloud inputs generated by multiple cameras fixed in the workspace. This type of point cloud input is not compatible with the now-common setting where the primary input modality is an eye-in-hand RGB camera like a GoPro. This paper closes this gap by incorporating into the diffusion policy model a process that projects features from the 2D RGB camera image onto a sphere. This enables us to reason about symmetries in SO(3) without explicitly reconstructing a point cloud. We perform extensive experiments in both simulation and the real world that demonstrate that our method consistently outperforms strong baselines in terms of both performance and sample efficiency. Our work, Image-to-Sphere Policy (ISP), is the first SO(3)-equivariant policy learning framework for robotic manipulation that works using only monocular RGB inputs.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Placed in Diffusion and flow policies for robotics, one of 14 directions in Diffusion models. Builds on 66 of its 76 references that are in the map. First posted on arXiv 22 May 2025.

Builds on

66 of 76 references.

Closest ancestors · 11

Same direction, newest first

  1. 2024ET-SEED: Efficient Trajectory-Level SE(3) Equivariant Diffusion Policy6 Nov
  2. 2024RDT-1B: a Diffusion Foundation Model for Bimanual Manipulation10 Oct
  3. 2024ActionFlow: Equivariant, Accurate, and Efficient Policies with Spatially Symmetric Flow Matching6 Sep
  4. 2024Diffusion Policy Policy Optimization1 Sep
  5. 2024Equivariant Diffusion Policy1 Jul
  6. 2024EquiBot: SIM(3)-Equivariant Diffusion Policy for Generalizable and Data Efficient Learning1 Jul
  7. 2024Consistency Policy: Accelerated Visuomotor Policies via Consistency Distillation13 May
  8. 2023Diffusion-EDFs: Bi-equivariant Denoising Generative Modeling on SE(3) for Visual Robotic Manipulation6 Sep
  9. 2023Goal-Conditioned Imitation Learning using Score-based Diffusion Policies5 Apr
  10. 2023EDGI: Equivariant Diffusion for Planning with Embodied Agents22 Mar
  11. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusion7 Mar

Other references · 55

  1. 2017Attention Is All You Needcited by 60,361
  2. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  3. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  4. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  5. 2020Denoising Diffusion Implicit Modelscited by 9,781
  6. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,498
  7. 2022RT-1: Robotics Transformer for Real-World Control at Scalecited by 2,061
  8. 2025π0.5: a Vision-Language-Action Model with Open-World Generalizationcited by 1,761
  9. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,568
  10. 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,214
  11. 2022Planning with Diffusion for Flexible Behavior Synthesiscited by 1,061
  12. 2015End-to-End Training of Deep Visuomotor Policiescited by 875
38 more
  1. 2023Open X-Embodiment: Robotic Learning Datasets and RT-X Modelscited by 862
  2. 2024DROID: A Large-Scale In-The-Wild Robot Manipulation Datasetcited by 852
  3. 2018Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point cloudscited by 746
  4. 2018QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulationcited by 743
  5. 2021What Matters in Learning from Offline Human Demonstrations for Robot Manipulationcited by 725
  6. 2025FAST: Efficient Action Tokenization for Vision-Language-Action Modelscited by 665
  7. 2019General E(2)-Equivariant Steerable CNNscited by 446
  8. 2024Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robotscited by 394
  9. 2023MimicGen: A Data Generation System for Scalable Robot Learning using Human Demonstrationscited by 327
  10. 2022Behavior Transformers: Cloning k modes with one stonecited by 304
  11. 2017Deep Imitation Learning for Complex Manipulation Tasks from Virtual Reality Teleoperationcited by 289
  12. 2022Equiformer: Equivariant Graph Attention Transformer for 3D Atomistic Graphscited by 264
  13. 2021MT-Opt: Continuous Multi-Task Robotic Reinforcement Learning at Scalecited by 229
  14. 2023TidyBot: Personalized Robot Assistance with Large Language Modelscited by 221
  15. 2021Neural Descriptor Fields: SE(3)-Equivariant Object Representations for Manipulationcited by 209
  16. 2023GELLO: A General, Low-Cost, and Intuitive Teleoperation Framework for Robot Manipulatorscited by 189
  17. 20233D Equivariant Diffusion for Target-Aware Molecule Generation and Affinity Predictioncited by 177
  18. 2024Data Scaling Laws in Imitation Learning for Robotic Manipulationcited by 145
  19. 2022SO(2)-Equivariant Reinforcement Learningcited by 84
  20. 2024UMI on Legs: Making Manipulation Policies Mobile with Manipulation-Centric Whole-body Controllerscited by 75
  21. 2021Equivariant Q Learning in Spatial Action Spacescited by 62
  22. 2022Sample Efficient Grasp Learning Using Equivariant Modelscited by 53
  23. 2022Look Closer: Bridging Egocentric and Third-Person Views with Transformers for Robotic Manipulationcited by 49
  24. 2022Equivariant Transporter Networkcited by 46
  25. 2022On-Robot Learning With Equivariant Modelscited by 42
  26. 2024FastUMI: A Scalable and Hardware-Independent Universal Manipulation Interface with Datasetcited by 40
  27. 2022The Surprising Effectiveness of Equivariant Models in Domains with Latent Symmetrycited by 33
  28. 2024Fourier Transporter: Bi-Equivariant Robotic Manipulation in 3Dcited by 33
  29. 2025BEHAVIOR Robot Suite: Streamlining Real-World Whole-Body Manipulation for Everyday Household Activitiescited by 31
  30. 2020Trajectory Prediction using Equivariant Continuous Convolutioncited by 26
  31. 2018Reinforcement Learning of Active Vision for Manipulating Objects under Occlusionscited by 25
  32. 2020Group Equivariant Deep Reinforcement Learningcited by 24
  33. 2022SEIL: Simulation-augmented Equivariant Imitation Learningcited by 23
  34. 2023Leveraging Symmetries in Pick and Placecited by 19
  35. 2023Image to Sphere: Learning Equivariant Features for Efficient Pose Predictioncited by 18
  36. 2024LEGATO: Cross-Embodiment Imitation Using a Grasping Toolcited by 18
  37. 2024OrbitGrasp: SE(3)-Equivariant Grasp Learningcited by 16
  38. 2025Two by Two: Learning Multi-Task Pairwise Objects Assembly for Generalizable Robot Manipulationcited by 13

50 most cited of 55 All 55 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 2; 2026: 3.

5 papers.

  1. 2025SE(3)-Equivariant Diffusion Policy in Spherical Fourier Spacecited by 8
  2. 2025Generalizable Hierarchical Skill Learning via Object-Centric Representationcited by 2
  3. 2026Action Map Policy: Learning 3D Closed-loop Manipulation via Pixel Classificationcited by 0
  4. 2026Pix2Act: Image-Space Manipulation Policies with Equivariant Augmentationcited by 0
  5. 2026Equivariant Visual-Tactile Diffusion Policy for Contact-Rich Manipulationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Efficient Hybrid SE(3)-Equivariant Visuomotor Flow Policy via Spherical Harmonics for Robot Manipulation
  2. 2025Canonical Policy: Learning Canonical 3D Representation for SE(3)-Equivariant Policy
  3. 2026EquiForm: Noise-Robust SE(3)-Equivariant Policy Learning from 3D Point Clouds
  4. 2025Spatial-Temporal Aware Visuomotor Diffusion Policy Learning
  5. 20253D Flow Diffusion Policy: Visuomotor Policy Learning via Generating Flow in 3D Space
  6. 2025ReSeFlow: Rectifying SE(3)-Equivariant Policy Learning Flows
  7. 2026PocketDP3: Efficient Pocket-Scale 3D Visuomotor Policy
  8. 2025A Practical Guide for Incorporating Symmetry in Diffusion Policy