mlsanityData to 7 Oct 2026

MSVIPER: Improved Policy Distillation for Reinforcement-Learning-Based Robot Navigation

arXiv 2209.09079 ↗First posted 19 Sep 2022cs.RO cs.AI cs.HCcited by 0

Not on the map. Why?

Abstract (arXiv)

We present Multiple Scenario Verifiable Reinforcement Learning via Policy Extraction (MSVIPER), a new method for policy distillation to decision trees for improved robot navigation. MSVIPER learns an "expert" policy using any Reinforcement Learning (RL) technique involving learning a state-action mapping and then uses imitation learning to learn a decision-tree policy from it. We demonstrate that MSVIPER results in efficient decision trees and can accurately mimic the behavior of the expert policy. Moreover, we present efficient policy distillation and tree-modification techniques that take advantage of the decision tree structure to allow improvements to a policy without retraining. We use our approach to improve the performance of RL-based robot navigation algorithms for indoor and outdoor scenes. We demonstrate the benefits in terms of reduced freezing and oscillation behaviors (by up to 95% reduction) for mobile robots navigating among dynamic obstacles and reduced vibrations and oscillation (by up to 17%) for outdoor robot navigation on complex, uneven terrains.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 33 of its 64 references that are in the map. First posted on arXiv 19 Sep 2022.

Builds on

33 of 64 references.

References · 33

  1. 2017Proximal Policy Optimization Algorithmscited by 14,441
  2. 2017Towards A Rigorous Science of Interpretable Machine Learningcited by 1,400
  3. 2017Constrained Policy Optimizationcited by 928
  4. 2019Learning by Cheatingcited by 361
  5. 2017Distilling a Neural Network Into a Soft Decision Treecited by 230
  6. 2018Verifiable Reinforcement Learning via Policy Extractioncited by 176
  7. 2019End-to-End Robotic Reinforcement Learning without Reward Engineeringcited by 115
  8. 2016Decentralized Non-communicating Multiagent Collision Avoidance with Deep Reinforcement Learningcited by 104
  9. 2017Uncertainty-Aware Reinforcement Learning for Collision Avoidancecited by 86
  10. 2017Towards Optimally Decentralized Multi-Robot Collision Avoidance via Deep Reinforcement Learningcited by 81
  11. 2020Cautious Adaptation For Reinforcement Learning in Safety-Critical Settingscited by 42
  12. 2019An Inductive Synthesis Framework for Verifiable Reinforcement Learningcited by 39
21 more
  1. 2019Conservative Q-Improvement: Reinforcement Learning for an Interpretable Decision-Tree Policycited by 29
  2. 2020Deep Reinforcement Learning based Automatic Exploration for Navigation in Unknown Environmentcited by 29
  3. 2019Risk-Aware Active Inverse Reinforcement Learningcited by 26
  4. 2020DenseCAvoid: Real-time Navigation in Dense Crowds using Anticipatory Behaviorscited by 25
  5. 2019DualSMC: Tunneling Differentiable Filtering and Planning under Continuous POMDPscited by 21
  6. 2021Vision-Based Autonomous Car Racing Using Deep Imitative Reinforcement Learningcited by 21
  7. 2020Risk-Aware High-level Decisions for Automated Driving at Occluded Intersections with Reinforcement Learningcited by 19
  8. 2020Frozone: Freezing-Free, Pedestrian-Friendly Navigation in Human Crowdscited by 14
  9. 2020Tactical Decision-Making in Autonomous Driving by Reinforcement Learning with Uncertainty Estimationcited by 14
  10. 2019Free-Lunch Saliency via Attention in Atari Agentscited by 12
  11. 2021Distillation of RL Policies with Formal Guarantees via Variational Abstraction of Markov Decision Processes (Technical Report)cited by 9
  12. 2018Getting Robots Unfrozen and Unlost in Dense Pedestrian Crowdscited by 5
  13. 2020OF-VO: Efficient Navigation among Pedestrians Using Commodity Sensorscited by 5
  14. 2021Neural-to-Tree Policy Distillation with Policy Improvement Criterioncited by 5
  15. 2018Towards Mixed Optimization for Reinforcement Learning with Program Synthesiscited by 4
  16. 2021XAI-N: Sensor-based Robot Navigation using Expert Policies and Decision Treescited by 3
  17. 2019Not All Features Are Equal: Feature Leveling Deep Neural Networks for Better Interpretationcited by 2
  18. 2020Improving Competence for Reliable Autonomycited by 2
  19. 2021Don't Get Yourself into Trouble! Risk-aware Decision-Making for Autonomous Vehiclescited by 2
  20. 2022Policy Distillation with Selective Input Gradient Regularization for Efficient Interpretabilitycited by 2
  21. 2019DynoPlan: Combining Motion Planning and Deep Neural Network based Controllers for Safe HRLcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2022Multi-Objective Policy Gradients with Topological Constraints
  2. 2018Reinforced Imitation: Sample Efficient Deep Reinforcement Learning for Map-less Navigation by Leveraging Prior Demonstrations
  3. 2017One-Shot Reinforcement Learning for Robot Navigation with Interactive Replay
  4. 2026VersualRL: Closed-Loop Verbal Reinforcement Learning with Visual Execution Feedback for Task-Level Robot Planning ↗
  5. 2026Beyond Imitation: Reinforcement Learning Fine-Tuning for Adaptive Diffusion Navigation Policies
  6. 2026Selective Agent Guidance via Entropy: Learning Autonomous Policies from Imperfect VLM Teachers
  7. 2022Offline Reinforcement Learning for Visual Navigation
  8. 2019Relay Policy Learning: Solving Long-Horizon Tasks via Imitation and Reinforcement Learning