mlsanityData to 7 Oct 2026

Learning Robotic Manipulation Policies from Point Clouds with Conditional Flow Matching

arXiv 2409.07343 ↗First posted 11 Sep 2024cs.ROcited by 63

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

also in 1 more
Abstract (arXiv)

Learning from expert demonstrations is a promising approach for training robotic manipulation policies from limited data. However, imitation learning algorithms require a number of design choices ranging from the input modality, training objective, and 6-DoF end-effector pose representation. Diffusion-based methods have gained popularity as they enable predicting long-horizon trajectories and handle multimodal action distributions. Recently, Conditional Flow Matching (CFM) (or Rectified Flow) has been proposed as a more flexible generalization of diffusion models. In this paper, we investigate the application of CFM in the context of robotic policy learning and specifically study the interplay with the other design choices required to build an imitation learning algorithm. We show that CFM gives the best performance when combined with point cloud input observations. Additionally, we study the feasibility of a CFM formulation on the SO(3) manifold and evaluate its suitability with a simplified example. We perform extensive experiments on RLBench which demonstrate that our proposed PointFlowMatch approach achieves a state-of-the-art average success rate of 67.8% over eight tasks, double the performance of the next best method.

Key facts. Cited by 63 papers in the map (data through 7 Oct 2026). Placed in Diffusion and flow policies for robotics, one of 14 directions in Diffusion models. Builds on 30 of its 33 references that are in the map. First posted on arXiv 11 Sep 2024.

Builds on

30 of 33 references.

Closest ancestors · 3

Same direction, newest first

  1. 2024Riemannian Flow Matching Policy for Robot Motion Learning15 Mar
  2. 2024Hierarchical Diffusion Policy for Kinematics-Aware Multi-Task Robotic Manipulation6 Mar
  3. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusion7 Mar

Other references · 27

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2020Denoising Diffusion Implicit Modelscited by 9,655
  3. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  4. 2022Flow Matching for Generative Modelingcited by 5,426
  5. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  6. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,094
  7. 2022Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flowcited by 3,550
  8. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,490
  9. 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,186
  10. 2022Planning with Diffusion for Flexible Behavior Synthesiscited by 1,023
  11. 2022Building Normalizing Flows with Stochastic Interpolantscited by 964
  12. 2019RLBench: The Robot Learning Benchmark & Learning Environmentcited by 669
15 more
  1. 2022Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulationcited by 545
  2. 2018An Algorithmic Perspective on Imitation Learningcited by 455
  3. 2021Implicit Behavioral Cloningcited by 442
  4. 2023Flow Matching on General Geometriescited by 260
  5. 2023SE(3)-Stochastic Flow Matching for Protein Backbone Generationcited by 147
  6. 2023Act3D: 3D Feature Field Transformers for Multi-Task Robotic Manipulationcited by 135
  7. 2018A micro Lie theory for state estimation in roboticscited by 117
  8. 2021Deep Regression on Manifolds: A 3D Rotation Case Studycited by 67
  9. 2022Interactive Imitation Learning in Robotics: A Surveycited by 51
  10. 2024AdaFlow: Imitation Learning with Variance-Adaptive Flow-Based Policiescited by 50
  11. 2024Learning with 3D rotations, a hitchhiker's guide to SO(3)cited by 39
  12. 2024Flow Matching Imitation Learning for Multi-Support Manipulationcited by 29
  13. 2024DITTO: Demonstration Imitation by Trajectory Transformationcited by 28
  14. 2021Correct Me if I am Wrong: Interactive Learning for Robotic Manipulationcited by 24
  15. 2023The Treachery of Images: Bayesian Scene Keypoints for Deep Policy Learning in Robotic Manipulationcited by 12

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 2; 2025: 33; 2026: 28.

63 papers.

Sort:

By area:

  1. 2025ManiFlow: A General Robot Manipulation Policy via Consistency Flow Trainingcited by 43
  2. 2025Much Ado About Noising: Dispelling the Myths of Generative Robotic Controlcited by 35
  3. 20253D FlowMatch Actor: Unified 3D Policy for Single- and Dual-Arm Manipulationcited by 19
  4. 2024Fast and Robust Visuomotor Riemannian Flow Matching Policycited by 18
  5. 2025Articulated Object Estimation in the Wildcited by 18
  6. 2025Chain-of-Action: Trajectory Autoregressive Modeling for Robotic Manipulationcited by 16
  7. 2025FreqPolicy: Efficient Flow-based Visuomotor Policy via Frequency Consistencycited by 13
  8. 2025FlowRAM: Grounding Flow Matching Policy with Region-Aware Mamba Framework for Robotic Manipulationcited by 13
  9. 2025Rethinking Bimanual Robotic Manipulation: Learning with Decoupled Interaction Frameworkcited by 11
  10. 2025SafeFlowMatcher: Safe and Fast Planning using Flow Matching with Control Barrier Functionscited by 11
  11. 2025PointMapPolicy: Structured Point Cloud Processing for Multi-Modal Imitation Learningcited by 10
  12. 2024FlowNav: Combining Flow Matching and Depth Priors for Efficient Navigationcited by 9
38 more
  1. 2025Conditioning Matters: Training Diffusion Policies is Faster Than You Thinkcited by 9
  2. 2025Fast Flow-based Visuomotor Policies via Conditional Optimal Transport Couplingscited by 8
  3. 2025DM1: MeanFlow with Dispersive Regularization for 1-Step Robotic Manipulationcited by 8
  4. 2025Uncertainty Comes for Free: Human-in-the-Loop Policies with Diffusion Modelscited by 6
  5. 2025Time-Unified Diffusion Policy with Action Discrimination for Robotic Manipulationcited by 5
  6. 2026One-Step Flow Policy: Self-Distillation for Fast Visuomotor Policiescited by 5
  7. 2025UniConFlow: A Unified Constrained Flow-Matching Framework for Certified Motion Planningcited by 4
  8. 2026From Flow to One Step: Real-Time Multi-Modal Trajectory Policies via Implicit Maximum Likelihood Estimation-based Distribution Distillationcited by 4
  9. 2026WarmPrior: Straightening Flow-Matching Policies with Temporal Priorscited by 4
  10. 2025Flow Matching Ergodic Coveragecited by 3
  11. 2025The Unreasonable Effectiveness of Discrete-Time Gaussian Process Mixtures for Robot Policy Learningcited by 3
  12. 2026Scaling Single Human Demonstrations for Imitation Learning using Generative Foundational Modelscited by 3
  13. 2025Designing for Difference: How Human Characteristics Shape Perceptions of Collaborative Robotscited by 2
  14. 2025Formal Safety Verification and Refinement for Generative Motion Planners via Certified Local Stabilizationcited by 2
  15. 2025Flow with the Force Field: Learning 3D Compliant Flow Matching Policies from Force and Demonstration-Guided Simulation Datacited by 2
  16. 2026Crowd-FM: Learned Optimal Selection of Conditional Flow Matching-generated Trajectories for Crowd Navigationcited by 2
  17. 2026Truncated Rectified Flow Policy for Reinforcement Learning with One-Step Samplingcited by 2
  18. 2026On the Generalization Capabilities, Design Choices and Limitations of Keypoint Imitation Learningcited by 2
  19. 2026Where Should Action Generation Begin? A Learnable Source Prior for Generative Robot Policiescited by 2
  20. 2025Gondola: Grounded Vision Language Planning for Robotic Manipulationcited by 1
  21. 2025Dynamic Objects Relocalization in Changing Environments with Flow Matchingcited by 1
  22. 2025MSG: Multi-Stream Generative Policies for Sample-Efficient Robotic Manipulationcited by 1
  23. 2025ResponsibleRobotBench: Benchmarking Responsible Robot Manipulation using Multi-modal Large Language Modelscited by 1
  24. 2026HyReach: Vision-Guided Hybrid Manipulator Reaching in Unseen Cluttered Environmentscited by 1
  25. 2026Jointly Learning Predicates and Actions Enables Zero-Shot Skill Compositioncited by 1
  26. 2026Implicit Drifting Policy: One-Step Action Generation via Conditional Expert Geometrycited by 1
  27. 2026Sample-efficient Low-level Motion Planning for Robotic Manipulation Tasks via Zero-shot Transfer Learningcited by 1
  28. 2026Invertible Neural Network Adapter for One-Step Flow Matching in Robot Manipulationcited by 1
  29. 2026High-Fidelity One-Step Generative Visuomotor Policy via Recursive Correction, Frequency Consistency, and Contrastive Flow Matchingcited by 1
  30. 2026Agile-WAM: An Agile Tactile World Action Model for Contact-Rich Robot Controlcited by 1
  31. 2025Composite Gaussian Processes Flows for Learning Discontinuous Multimodal Policiescited by 0
  32. 2025GIFT: Geometry-Induced Functional Transfer for Category-level Object Manipulationcited by 0
  33. 2025Unified Generation-Refinement Planning: Bridging Guided Flow Matching and Sampling-Based MPC for Social Navigationcited by 0
  34. 2025Towards Safe Imitation Learning via Potential Field-Guided Flow Matchingcited by 0
  35. 2025Fine-Tuning Flow Matching via Maximum Likelihood Estimation of Reconstructionscited by 0
  36. 2025Learning Generalizable Visuomotor Policy through Dynamics-Alignmentcited by 0
  37. 2025Structured Imitation Learning of Interactive Policies through Inverse Gamescited by 0
  38. 2025Kinematics-Aware Diffusion Policy with Consistent 3D Observation and Action Space for Whole-Arm Robotic Manipulationcited by 0

50 most cited of 63 All 63 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2024FlowPolicy: Enabling Fast and Robust 3D Flow-based Policy via Consistency Flow Matching for Robot Manipulation
  2. 2022ToolFlowNet: Robotic Manipulation with Tools via Predicting Tool Flow from Point Clouds
  3. 2026Flowing With Purpose: Latent Action Guided Flow Matching Policies For Robotic Manipulation
  4. 2025MP1: MeanFlow Tames Policy Learning in 1-step for Robotic Manipulation
  5. 2025VO-DP: Semantic-Geometric Adaptive Diffusion Policy for Vision-Only Robotic Manipulation
  6. 2026BayesFP: Posterior Estimation for Flow-Based Policies via Feynman-Kac Sampling
  7. 2025ISS Policy : Scalable Diffusion Policy with Implicit Scene Supervision
  8. 2025Streaming Flow Policy: Simplifying diffusion/flow-matching policies by treating action trajectories as flow trajectories