TransCurriculum: Multi-Dimensional Curriculum Learning for Fast & Stable Locomotion

• URL: http://arxiv.org/abs/2603.14156v1
• Date: Sat, 14 Mar 2026 23:44:42 GMT
• Title: TransCurriculum: Multi-Dimensional Curriculum Learning for Fast & Stable Locomotion
• Abstract summary: TransCurriculum is a transformer-based multi-dimensional curriculum learning approach for agile quadrupedal locomotion.<n>We validate our approach on the Unitree Go1 robot in simulation and deploy it zero-shot on Go1 hardware.
• Score: 50.54752207285298
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: High-speed legged locomotion struggles with stability and transfer losses at higher command velocities during deployment. One reason is that most curricula vary difficulty along single axis, for example increase the range of command velocities, terrain difficulty, or domain parameters (e.g. friction or payload mass) using either fixed update rule or instantaneous rewards while ignoring how the history of robot training has evolved. We propose TransCurriculum, a transformer-based multi-dimensional curriculum learning approach for agile quadrupedal locomotion. TransCurriculum adapts to 3 axes, velocity command targets, terrain difficulty, and domain randomization parameters (friction and payload mass). Rather than feeding task reward history directly into the low-level control policy, our formulation exploits it at the curriculum level. A transformer-based teacher retrieves the sequence of rewards and uses it to predict future rewards, success rate, and learning progress to guide expansion of this multidimensional curriculum towards high performing task bins. Finally we validate our approach on the Unitree Go1 robot in simulation (Isaac Gym) and deploy it zero-shot on Go1 hardware. Our TransCurriculum policy achieves a maximum velocity of 6.3 m/s in simulation and outperforms prior curriculum baselines. We tested our TransCurriculum trained policy on terrains (carpets, slopes, tiles, concrete), achieving a forward velocity of 4.1 m/s on carpet surpassing the fastest curriculum methods by 18.8% and achieves maximum zero-shot value among all tested methods. Our multi-dimensional curriculum also reduces the transfer loss to 18% from 27% for command only curriculum, demonstrating the benefits of joint training over velocity, terrain and domain randomization dimension while keeping the task success rate of 80-90% on rigid indoor and outdoor surfaces.

Related papers

• Scaling Curriculum Learning For Autonomous Driving [35.25765047709282]
  We present CL4AD, the first integration of curriculum learning into batched autonomous driving simulators.<n>We introduce utility functions that shape curricula based on success rates and the realism of the agent's behavior.<n>We show that curriculum learning achieves a 99% success rate a billion steps earlier than domain randomization.
  arXiv  Detail & Related papers  (2026-08-23T18:51:05Z)
• Simple but Stable, Fast and Safe: Achieve End-to-end Control by High-Fidelity Differentiable Simulation [14.322775509774246]
  Obstacle avoidance is a fundamental vision-based task essential for enabling quadrotors to perform advanced applications.<n>We propose a novel end-to-end policy that directly maps depth images to low-level bodyrate commands by reinforcement learning.<n>Results demonstrate that the proposed approach achieves the highest success rate and the lowest jerk among state-of-the-art baselines.
  arXiv  Detail & Related papers  (2026-04-12T09:38:00Z)
• APEX: Learning Adaptive High-Platform Traversal for Humanoid Robots [32.84976822343718]
  We propose APEX, a system for perceptive, climbing-based high-platform traversal behaviors.<n>Central to our approach is a generalized ratchet progress reward for learning contact-rich, goal-reaching maneuvers.<n>We train LiDAR-based full-body maneuver policies and reduce the sim-to-real perception gap.
  arXiv  Detail & Related papers  (2026-02-11T18:55:11Z)
• Gait-Adaptive Perceptive Humanoid Locomotion with Real-Time Under-Base Terrain Reconstruction [86.53248703859718]
  We introduce a perceptive locomotion framework that merges terrain sensing, gait regulation, and whole-body control into a single reinforcement learning policy.<n>Experiments conducted on a 31-DoF, 1.65 m humanoid robot demonstrate robust locomotion in both simulation and real-world settings.
  arXiv  Detail & Related papers  (2025-12-08T11:42:41Z)
• Learning Terrain-Specialized Policies for Adaptive Locomotion in Challenging Environments [0.4588028371034407]
  This work introduces a hierarchical reinforcement learning framework that leverages terrain-specialized policies and curriculum learning to enhance agility and tracking performance in complex environments.<n>We validated our method on simulation, where our approach outperforms a generalist policy by up to 16% in success rate and achieves lower tracking errors as the velocity target increases.
  arXiv  Detail & Related papers  (2025-09-25T00:17:39Z)
• Single-Shot Learning of Stable Dynamical Systems for Long-Horizon Manipulation Tasks [48.54757719504994]
  This paper focuses on improving task success rates while reducing the amount of training data needed. Our approach introduces a novel method that segments long-horizon demonstrations into discrete steps defined by waypoints and subgoals. We validate our approach through both simulation and real-world experiments, demonstrating effective transfer from simulation to physical robotic platforms.
  arXiv  Detail & Related papers  (2024-10-01T19:49:56Z)
• FLaRe: Achieving Masterful and Adaptive Robot Policies with Large-Scale Reinforcement Learning Fine-Tuning [74.25049012472502]
  FLaRe is a large-scale Reinforcement Learning framework that integrates robust pre-trained representations, large-scale training, and gradient stabilization techniques. Our method aligns pre-trained policies towards task completion, achieving state-of-the-art (SoTA) performance on previously demonstrated and on entirely novel tasks and embodiments.
  arXiv  Detail & Related papers  (2024-09-25T03:15:17Z)
• Autonomous Vehicle Controllers From End-to-End Differentiable Simulation [57.278726604424556]
  We propose a differentiable simulator and design an analytic policy gradients (APG) approach to training AV controllers.<n>Our proposed framework brings the differentiable simulator into an end-to-end training loop, where gradients of environment dynamics serve as a useful prior to help the agent learn a more grounded policy.<n>We find significant improvements in performance and robustness to noise in the dynamics, as well as overall more intuitive human-like handling.
  arXiv  Detail & Related papers  (2024-09-12T11:50:06Z)
• Rethinking Closed-loop Training for Autonomous Driving [82.61418945804544]
  We present the first empirical study which analyzes the effects of different training benchmark designs on the success of learning agents. We propose trajectory value learning (TRAVL), an RL-based driving agent that performs planning with multistep look-ahead. Our experiments show that TRAVL can learn much faster and produce safer maneuvers compared to all the baselines.
  arXiv  Detail & Related papers  (2023-06-27T17:58:39Z)
• Advanced Skills by Learning Locomotion and Local Navigation End-to-End [10.872193480485596]
  In this work, we propose to solve the complete problem by training an end-to-end policy with deep reinforcement learning. We demonstrate the successful deployment of policies on a real quadrupedal robot.
  arXiv  Detail & Related papers  (2022-09-26T16:35:00Z)
• Rapid Locomotion via Reinforcement Learning [15.373208553045416]
  We present an end-to-end learned controller that achieves record agility for the MIT Mini Cheetah. This system runs and turns fast on natural terrains like grass, ice, and gravel and responds robustly to disturbances.
  arXiv  Detail & Related papers  (2022-05-05T17:55:11Z)
• Real Robot Challenge using Deep Reinforcement Learning [6.332038240397164]
  This paper details our winning submission to Phase 1 of the 2021 Real Robot Challenge. The challenge is in which a three fingered robot must carry a cube along specified goal trajectories. We use a pure reinforcement learning approach which requires minimal expert knowledge of the robotic system.
  arXiv  Detail & Related papers  (2021-09-30T16:12:17Z)
• Robust High-speed Running for Quadruped Robots via Deep Reinforcement Learning [7.264355680723856]
  In this paper, we explore learning foot positions in Cartesian space for a task of running as fast as possible subject to environmental disturbances. Compared with other action spaces, we observe less needed reward shaping, much improved sample efficiency, and the emergence of natural gaits such as galloping and bounding. Policies can be learned in only a few million time steps, even for challenging tasks of running over rough terrain with loads of over 100% of the nominal quadruped mass.
  arXiv  Detail & Related papers  (2021-03-11T06:13:09Z)

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.