Simulator Adaptation for Sim-to-Real Learning of Legged Locomotion via Proprioceptive Distribution Matching

• URL: http://arxiv.org/abs/2604.11090v1
• Date: Mon, 13 Apr 2026 07:10:52 GMT
• Title: Simulator Adaptation for Sim-to-Real Learning of Legged Locomotion via Proprioceptive Distribution Matching
• Abstract summary: Simulation trained legged locomotion policies exhibit performance loss on hardware due to dynamics discrepancies between the simulator and the real world.<n>Prior work typically quantify these discrepancies through precise, time-aligned matching of joint and base trajectories.<n>We introduce a practical alternative based on proprioceptive distribution matching, which compares hardware and simulation rollouts as distributions of joint observations and actions.
• Score: 16.52598063527788
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Simulation trained legged locomotion policies often exhibit performance loss on hardware due to dynamics discrepancies between the simulator and the real world, highlighting the need for approaches that adapt the simulator itself to better match hardware behavior. Prior work typically quantify these discrepancies through precise, time-aligned matching of joint and base trajectories. This process requires motion capture, privileged sensing, and carefully controlled initial conditions. We introduce a practical alternative based on proprioceptive distribution matching, which compares hardware and simulation rollouts as distributions of joint observations and actions, eliminating the need for time alignment or external sensing. Using this metric as a black-box objective, we explore adapting simulator dynamics through parameter identification, action-delta models, and residual actuator models. Our approach matches the parameter recovery and policy-performance gains of privileged state-matching baselines across extensive sim-to-sim ablations on the Go2 quadruped. Real-world experiments demonstrate substantial drift reduction using less than five minutes of hardware data, even for a challenging two-legged walking behavior. These results demonstrate that proprioceptive distribution matching provides a practical and effective route to simulator adaptation for sim-to-real transfer of learned legged locomotion.

Related papers

• Pre-training Visual Dexterity in Simulation [67.17477115017896]
  We introduce a pre-training framework for dexterous manipulation that uses data entirely collected in simulation.<n>With the help of five operators, we collect 75 hours of multi-task dexterous manipulation over one week, and use it to pre-train a causal transformer on a sequence modeling objective.<n>We find that our approach outperforms training behavior cloning policies from scratch, showing that simulation teleoperation is a viable pre-training source for real-world dexterous manipulation.
  arXiv  Detail & Related papers  (2026-08-16T20:21:33Z)
• WorldSimProbe: Diagnosing Simulator Faithfulness in Action-Conditioned World Models for Embodied Manipulation [85.15806404962193]
  Action-conditioned world models (ACWMs) promise to provide embodied AI with scalable predictive simulators for planning, policy evaluation, and data generation.<n>We evaluate ACWMs through the observable capabilities expected of physical simulators.<n>WorldSimProbe reveals systematic action-realization degradation across control variation, structured failures in interaction grounding and dynamics.
  arXiv  Detail & Related papers  (2026-08-10T08:48:06Z)
• TactSpace: Learning a Physics-enriched Shared Latent Space for Tactile Sim-to-Real Transfer [8.357674069719524]
  We propose a multi-modal representation learning framework that aligns heterogeneous tactile modalities within a shared latent space.<n>Our approach employs modality-specific encoders to project diverse tactile observations into a common embedding space.<n>Our results demonstrate zero-shot sim-to-real transfer across physically dissimilar representations.
  arXiv  Detail & Related papers  (2026-06-17T11:41:27Z)
• RLFTSim: Realistic and Controllable Multi-Agent Traffic Simulation via Reinforcement Learning Fine-Tuning [19.985714275376726]
  We introduce RLFTSim, a reinforcement-learning-based fine-tuning framework.<n>It enhances scenario realism by aligning simulator rollouts with real-world data distributions.<n>We also demonstrate the effectiveness of our approach for distilling traffic simulation controllability through goal conditioning.
  arXiv  Detail & Related papers  (2026-05-18T18:57:13Z)
• Toward Efficient and Robust Behavior Models for Multi-Agent Driving Simulation [4.713158796503602]
  Multi-agent driving simulation requires realistic and computationally efficient behavior models.<n>We adopt an instance-centric scene representation, where each traffic participant and element map is modeled in its own local coordinate frame.<n>To model interactions, we employ a query-centric symmetric context encoder with relative positional encodings between local frames.
  arXiv  Detail & Related papers  (2025-12-05T15:32:36Z)
• DiffSim2Real: Deploying Quadrupedal Locomotion Policies Purely Trained in Differentiable Simulation [35.76143996968696]
  We show that locomotion policies trained with analytic gradients from a differentiable simulator can be successfully transferred to the real world. A key factor in our success is a smooth contact model that combines informative gradients with physical accuracy. This is the first time a real quadpedal robot is able to locomote after training exclusively in a differentiable simulation.
  arXiv  Detail & Related papers  (2024-11-04T15:43:57Z)
• Bridging the Sim-to-Real Gap with Bayesian Inference [53.61496586090384]
  We present SIM-FSVGD for learning robot dynamics from data. We use low-fidelity physical priors to regularize the training of neural network models. We demonstrate the effectiveness of SIM-FSVGD in bridging the sim-to-real gap on a high-performance RC racecar system.
  arXiv  Detail & Related papers  (2024-03-25T11:29:32Z)
• Learning Quadruped Locomotion Using Differentiable Simulation [31.80380408663424]
  Differentiable simulation promises fast convergence and stable training. This work proposes a new differentiable simulation framework to overcome these challenges. Our framework enables learning quadruped walking in simulation in minutes without parallelization.
  arXiv  Detail & Related papers  (2024-03-21T22:18:59Z)
• Waymax: An Accelerated, Data-Driven Simulator for Large-Scale Autonomous Driving Research [76.93956925360638]
  Waymax is a new data-driven simulator for autonomous driving in multi-agent scenes. It runs entirely on hardware accelerators such as TPUs/GPUs and supports in-graph simulation for training. We benchmark a suite of popular imitation and reinforcement learning algorithms with ablation studies on different design decisions.
  arXiv  Detail & Related papers  (2023-10-12T20:49:15Z)
• Perception Imitation: Towards Synthesis-free Simulator for Autonomous Vehicles [45.27200446670184]
  We propose a perception imitation method to simulate results of a certain perception model, and discuss a new route of autonomous driving simulator without data synthesis. Experiments show that our method is effective to model the behavior of learning-based perception model, and can be further applied in the proposed simulation route smoothly.
  arXiv  Detail & Related papers  (2023-04-19T01:27:02Z)
• DriveGAN: Towards a Controllable High-Quality Neural Simulation [147.6822288981004]
  We introduce a novel high-quality neural simulator referred to as DriveGAN. DriveGAN achieves controllability by disentangling different components without supervision. We train DriveGAN on multiple datasets, including 160 hours of real-world driving data.
  arXiv  Detail & Related papers  (2021-04-30T15:30:05Z)
• TrafficSim: Learning to Simulate Realistic Multi-Agent Behaviors [74.67698916175614]
  We propose TrafficSim, a multi-agent behavior model for realistic traffic simulation. In particular, we leverage an implicit latent variable model to parameterize a joint actor policy. We show TrafficSim generates significantly more realistic and diverse traffic scenarios as compared to a diverse set of baselines.
  arXiv  Detail & Related papers  (2021-01-17T00:29:30Z)
• Point Cloud Based Reinforcement Learning for Sim-to-Real and Partial Observability in Visual Navigation [62.22058066456076]
  Reinforcement Learning (RL) represents powerful tools to solve complex robotic tasks. RL does not work directly in the real-world, which is known as the sim-to-real transfer problem. We propose a method that learns on an observation space constructed by point clouds and environment randomization.
  arXiv  Detail & Related papers  (2020-07-27T17:46:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.