mlsanityData to 14 Oct 2026

Dream to Generalize: Zero-Shot Model-Based Reinforcement Learning for Unseen Visual Distractions

arXiv 2506.05419 ↗First posted 5 Jun 2025cs.CV cs.AIcited by 4

On the map: World models → Latent dynamics models and model-based RL

also in 1 more
Abstract (arXiv)

Model-based reinforcement learning (MBRL) has been used to efficiently solve vision-based control tasks in highdimensional image observations. Although recent MBRL algorithms perform well in trained observations, they fail when faced with visual distractions in observations. These task-irrelevant distractions (e.g., clouds, shadows, and light) may be constantly present in real-world scenarios. In this study, we propose a novel self-supervised method, Dream to Generalize (Dr. G), for zero-shot MBRL. Dr. G trains its encoder and world model with dual contrastive learning which efficiently captures task-relevant features among multi-view data augmentations. We also introduce a recurrent state inverse dynamics model that helps the world model to better understand the temporal structure. The proposed methods can enhance the robustness of the world model against visual distractions. To evaluate the generalization performance, we first train Dr. G on simple backgrounds and then test it on complex natural video backgrounds in the DeepMind Control suite, and the randomizing environments in Robosuite. Dr. G yields a performance improvement of 117% and 14% over prior works, respectively. Our code is open-sourced and available at https://github.com/JeongsooHa/DrG.git

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Placed in Latent dynamics models and model-based RL, one of 11 directions in World models. Builds on 29 of its 36 references that are in the map. First posted on arXiv 5 Jun 2025.

Builds on

29 of 36 references.

Closest ancestors · 3

Same direction, newest first

  1. 2021DreamerPro: Reconstruction-Free Model-Based Reinforcement Learning with Prototypical Representations27 Oct
  2. 2020Mastering Atari with Discrete World Models5 Oct
  3. 2019Dream to Control: Learning Behaviors by Latent Imagination3 Dec

Other references · 26

  1. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
  2. 2018Representation Learning with Contrastive Predictive Codingcited by 7,777
  3. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
  4. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,750
  5. 2015Continuous control with deep reinforcement learningcited by 3,407
  6. 2013Playing Atari with Deep Reinforcement Learningcited by 3,274
  7. 2015High-Dimensional Continuous Control Using Generalized Advantage Estimationcited by 2,026
  8. 2017The Kinetics Human Action Video Datasetcited by 1,909
  9. 2018Unsupervised Feature Learning via Non-Parametric Instance-level Discriminationcited by 1,845
  10. 2014On the Properties of Neural Machine Translation: Encoder-Decoder Approachescited by 1,304
  11. 2018Learning Latent Dynamics for Planning from Pixelscited by 1,264
  12. 2018Soft Actor-Critic Algorithms and Applicationscited by 1,240
14 more
  1. 2018DeepMind Control Suitecited by 956
  2. 2015End-to-End Training of Deep Visuomotor Policiescited by 875
  3. 2020CURL: Contrastive Unsupervised Representations for Reinforcement Learningcited by 760
  4. 2018QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulationcited by 743
  5. 2018Learning Dexterous In-Hand Manipulationcited by 687
  6. 2019Data-Efficient Image Recognition with Contrastive Predictive Codingcited by 673
  7. 2020Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixelscited by 585
  8. 2020robosuite: A Modular Simulation Framework and Benchmark for Robot Learningcited by 511
  9. 2018Visual Reinforcement Learning with Imagined Goalscited by 251
  10. 2020Generalization in Reinforcement Learning by Soft Data Augmentationcited by 137
  11. 2018Deep Variational Reinforcement Learning for POMDPscited by 125
  12. 2020Self-Supervised Policy Adaptation during Deploymentcited by 123
  13. 2021Temporal Predictive Coding For Model-Based Planning In Latent Spacecited by 55
  14. 2021SECANT: Self-Expert Cloning for Zero-Shot Generalization of Visual Policiescited by 43

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 3.

4 papers.

  1. 2026World Action Models: The Next Frontier in Embodied AIcited by 50
  2. 2026Aligning Agentic World Models via Knowledgeable Experience Learningcited by 5
  3. 2025DeGuV: Depth-Guided Visual Reinforcement Learning for Generalization and Interpretability in Manipulationcited by 0
  4. 2026VIGOR: Zero-Shot Visual Generalization via Latent-Space Consistency in Model-Based Reinforcement Learningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Make the Pertinent Salient: Task-Relevant Reconstruction for Visual Control with Distractions
  2. 2024Focus On What Matters: Separated Models For Visual-Based RL Generalization
  3. 2025Self-Predictive Dynamics for Generalization of Vision-based Reinforcement Learning
  4. 2024Zero-Shot Generalization of Vision-Based RL Without Data Augmentation
  5. 2024Dreaming of Many Worlds: Learning Contextual World Models Aids Zero-Shot Generalization
  6. 2025Self-Consistent Model-based Adaptation for Visual Reinforcement Learning
  7. 2025Pre-trained Visual Representations Generalize Where it Matters in Model-Based Reinforcement Learning
  8. 2025Synthetic Data is Sufficient for Zero-Shot Visual Generalization from Offline Data