# ObstaDiff: Generalizable Diffusion Policy Learning via Obstacle-aware Representations

arXiv: https://arxiv.org/abs/2609.10918 · HTML: https://mlsanity.com/paper/2609.10918/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Diffusion and flow policies for robotics, one of 14 directions in Diffusion models.
- Builds on 27 of its 32 references that are in the map.
- First posted on arXiv 10 Sep 2026.

## Placed in

- Diffusion models / [Diffusion and flow policies for robotics](https://mlsanity.com/diffusion/diffusion-and-flow-policies-for-robotics/)
- Offline RL and imitation learning from data / [Robot manipulation imitation learning](https://mlsanity.com/offlinerl/robot-manipulation-imitation-learning/)

## Builds on (most cited first, 27 of 27)

- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [Denoising Diffusion Implicit Models](https://mlsanity.com/paper/2010.02502/) (2020)
- [Diffusion Policy: Visuomotor Policy Learning via Action Diffusion](https://mlsanity.com/paper/2303.04137/) (2023)
- [Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware](https://mlsanity.com/paper/2304.13705/) (2023)
- [Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World](https://mlsanity.com/paper/1703.06907/) (2017)
- [Planning with Diffusion for Flexible Behavior Synthesis](https://mlsanity.com/paper/2205.09991/) (2022)
- [R3M: A Universal Visual Representation for Robot Manipulation](https://mlsanity.com/paper/2203.12601/) (2022)
- [YOLO-World: Real-Time Open-Vocabulary Object Detection](https://mlsanity.com/paper/2401.17270/) (2024)
- [Implicit Behavioral Cloning](https://mlsanity.com/paper/2109.00137/) (2021)
- [Faster Segment Anything: Towards Lightweight SAM for Mobile Applications](https://mlsanity.com/paper/2306.14289/) (2023)
- [Transporter Networks: Rearranging the Visual World for Robotic Manipulation](https://mlsanity.com/paper/2010.14406/) (2020)
- [3D Diffuser Actor: Policy Diffusion with 3D Scene Representations](https://mlsanity.com/paper/2402.10885/) (2024)
- [Real-World Robot Learning with Masked Visual Pre-training](https://mlsanity.com/paper/2210.03109/) (2022)
- [Motion Planning Diffusion: Learning and Planning of Robot Motions with Diffusion Models](https://mlsanity.com/paper/2308.01557/) (2023)
- [EquiBot: SIM(3)-Equivariant Diffusion Policy for Generalizable and Data Efficient Learning](https://mlsanity.com/paper/2407.01479/) (2024)
- [Motion Planning Networks: Bridging the Gap Between Learning-based and Classical Motion Planners](https://mlsanity.com/paper/1907.06013/) (2019)
- [Equivariant Diffusion Policy](https://mlsanity.com/paper/2407.01812/) (2024)
- [Learning Generalizable Manipulation Policies with Object-Centric 3D Representations](https://mlsanity.com/paper/2310.14386/) (2023)
- [Sparse Diffusion Policy: A Sparse, Reusable, and Flexible Policy for Robot Learning](https://mlsanity.com/paper/2407.01531/) (2024)
- [Motion Policy Networks](https://mlsanity.com/paper/2210.12209/) (2022)
- [What Makes Pre-Trained Visual Representations Successful for Robust Manipulation?](https://mlsanity.com/paper/2312.12444/) (2023)
- [Learned Visual Navigation for Under-Canopy Agricultural Robots](https://mlsanity.com/paper/2107.02792/) (2021)
- [Language-Guided Object-Centric Diffusion Policy for Generalizable and Collision-Aware Robotic Manipulation](https://mlsanity.com/paper/2407.00451/) (2024)
- [Augmented Reality for RObots (ARRO): Pointing Visuomotor Policies Towards Visual Robustness](https://mlsanity.com/paper/2505.08627/) (2025)
- [Autonomous Sweet Pepper Harvesting for Protected Cropping Systems](https://mlsanity.com/paper/1706.02023/) (2017)
- [Robust Visual Sim-to-Real Transfer for Robotic Manipulation](https://mlsanity.com/paper/2307.15320/) (2023)
- [EHC-MM: Embodied Holistic Control for Mobile Manipulation](https://mlsanity.com/paper/2409.08527/) (2024)

## Abstract (arXiv)

Imitation learning has achieved impressive results in robotic manipulation, yet most existing approaches assume clean backgrounds and lack explicit mechanisms for obstacle-aware motion generation. Extending such policies to cluttered, real-world scenes with unstructured obstacles remains a key generalization challenge. We present ObstaDiff, a decomposed diffusion-policy framework with a lightweight obstacle-aware visual encoder. ObstaDiff extracts a structured target-obstacle-background representation, enabling the downstream alignment policy to generate end-effector trajectories toward a target-centered bottleneck pose while reasoning about surrounding obstacles. We evaluate ObstaDiff on 61 real-robot greenhouse trials per method (366 executions in total). ObstaDiff achieves 75.41% average task success and 8.20% average obstacle collision rate, outperforming representative imitation-learning baselines and improving generalization in cluttered agricultural scenes.