mlsanityData to 14 Oct 2026

Distributionally Robust Policy Learning via Adversarial Environment Generation

arXiv 2107.06353 ↗First posted 13 Jul 2021cs.RO cs.LGcited by 17

Not on the map. Why?

Abstract (arXiv)

Our goal is to train control policies that generalize well to unseen environments. Inspired by the Distributionally Robust Optimization (DRO) framework, we propose DRAGEN - Distributionally Robust policy learning via Adversarial Generation of ENvironments - for iteratively improving robustness of policies to realistic distribution shifts by generating adversarial environments. The key idea is to learn a generative model for environments whose latent variables capture cost-predictive and realistic variations in environments. We perform DRO with respect to a Wasserstein ball around the empirical distribution of environments by generating realistic adversarial environments via gradient ascent on the latent space. We demonstrate strong Out-of-Distribution (OoD) generalization in simulation for (i) swinging up a pendulum with onboard vision and (ii) grasping realistic 3D objects. Grasping experiments on hardware demonstrate better sim2real performance compared to domain randomization.

Key facts. Cited by 17 papers in the map (data through 14 Oct 2026). Builds on 29 of its 48 references that are in the map. First posted on arXiv 13 Jul 2021.

Builds on

29 of 48 references.

References · 29

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2014Explaining and Harnessing Adversarial Examplescited by 7,364
  3. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 7,005
  4. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
  5. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,248
  6. 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,885
  7. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,897
  8. 2019DeepSDF: Learning Continuous Signed Distance Functions for Shape Representationcited by 2,689
  9. 2017Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real Worldcited by 1,498
  10. 2020Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixelscited by 585
  11. 2020Reinforcement Learning with Augmented Datacited by 468
  12. 2018Generalizing to Unseen Domains via Adversarial Data Augmentationcited by 451
17 more
  1. 2015Data-driven Distributionally Robust Optimization Using the Wasserstein Metric: Performance Guarantees and Tractable Reformulationscited by 291
  2. 2020Emergent Complexity and Zero-shot Transfer via Unsupervised Environment Designcited by 275
  3. 2019Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutionscited by 192
  4. 2016Distributionally Robust Stochastic Optimization with Wasserstein Distancecited by 185
  5. 2019Active Domain Randomizationcited by 170
  6. 2018Learning Synergies between Pushing and Grasping with Self-supervised Deep Reinforcement Learningcited by 128
  7. 2020Enhanced POET: Open-Ended Reinforcement Learning through Unbounded Invention of Learning Challenges and their Solutionscited by 106
  8. 2018The Limits and Potentials of Deep Learning for Roboticscited by 104
  9. 2020EGAD! an Evolved Grasping Analysis Dataset for diversity and reproducibility in robotic manipulationcited by 61
  10. 2019Semantic Adversarial Attacks: Parametric Transformations That Fool Deep Classifierscited by 57
  11. 2020Learning perturbation sets for robust machine learningcited by 48
  12. 2020Adversarial Generation of Continuous Implicit Shape Representationscited by 46
  13. 2019DeceptionNet: Network-Driven Domain Randomizationcited by 43
  14. 2017Domain Randomization and Generative Models for Robotic Graspingcited by 36
  15. 2017The Robust Manifold Defense: Adversarial Training using Generative Modelscited by 35
  16. 2020Model-Based Robust Deep Learning: Generalizing to Natural, Out-of-Distribution Datacited by 27
  17. 2020ShapeAdv: Generating Shape-Aware Adversarial 3D Point Cloudscited by 20

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 2; 2022: 0; 2023: 3; 2024: 3; 2025: 2; 2026: 7.

17 papers.

Sort:

By area:

  1. 2023AdaptSim: Task-Driven Simulation Adaptation for Sim-to-Real Transfercited by 18
  2. 2024Adversarial Environment Design via Regret-Guided Diffusion Modelscited by 10
  3. 2024RoboMD: Uncovering Robot Vulnerabilities through Semantic Potential Fieldscited by 7
  4. 2025Distributionally Robust Optimization via Diffusion Ambiguity Modelingcited by 4
  5. 2026Distributionally Robust Optimization via Generative Ambiguity Modelingcited by 3
  6. 2023Wasserstein Distributionally Robust Control Barrier Function using Conditional Value-at-Risk with Differentiable Convex Programmingcited by 2
  7. 2024Distributionally Robust Policy and Lyapunov-Certificate Learningcited by 2
  8. 2025Distributionally Robust Multi-Agent Reinforcement Learning for Dynamic Chute Mappingcited by 2
  9. 2021Learning Provably Robust Motion Planners Using Funnel Librariescited by 1
  10. 2021Stronger Generalization Guarantees for Robot Learning by Combining Generative Models and Real-World Datacited by 1
  11. 2023Distributionally Safe Reinforcement Learning under Model Uncertainty: A Single-Level Approach by Differentiable Convex Programmingcited by 1
  12. 2026Reward-Preserving Attacks For Robust Reinforcement Learningcited by 0
5 more
  1. 20263D-Learning: Diffusion-Augmented Distributionally Robust Decision-Focused Learningcited by 0
  2. 2026Taming the Curses of Multiagency in Robust Markov Games with Large State Space through Linear Function Approximationcited by 0
  3. 2026Generative Distributionally Robust Optimizationcited by 0
  4. 2026Curriculum Generation under Structured Parametric Environments for Robust Navigation Policiescited by 0
  5. 2026A Unified Perspective on Conformal Prediction and Wasserstein Distributionally Robust Optimization for Uncertainty Quantificationcited by 0