This repository is a curated list of resources for legged locomotion learning in robotics, targeting researchers and engineers in the field. It provides a comprehensive collection of papers, code repositories, and robot models to accelerate development and understanding of advanced locomotion techniques.

How It Works

The list is organized by robot type (bipeds, humanoids, quadrupeds) and includes links to relevant codebases, simulation environments, and research papers. It highlights key aspects like available robot models, simulation formats (URDF, MJCF), and licensing information, facilitating quick assessment of resources for specific projects.

Quick Start & Requirements

This is a curated list, not a runnable codebase. The linked resources may have their own installation and dependency requirements, often including Python, deep learning frameworks (PyTorch, TensorFlow), simulation environments (Isaac Gym, MuJoCo, PyBullet, Gazebo), and specific hardware like GPUs.

Highlighted Details

• Extensive collection of robot models with details on formats, licenses, and availability of meshes, inertias, and collision data.
• Links to numerous code repositories implementing various locomotion learning techniques, including reinforcement learning, imitation learning, and model-predictive control.
• Comprehensive lists of survey papers and recent publications from top robotics conferences (ICRA, IROS, CoRL, RSS) and journals.
• Includes resources for sim-to-real transfer, domain randomization, and specific robot platforms like Cassie, ANYmal, and Unitree robots.

Maintenance & Community

The list is maintained by gaiyi7788 and is updated gradually. Related "awesome" lists are linked, suggesting a broader community effort in curating robotics resources.

Licensing & Compatibility

Licenses vary significantly across the linked resources, ranging from permissive licenses like MIT and Apache-2.0 to more restrictive ones like GPL-3.0 and CC-BY-SA-4.0. Some robot models, like Boston Dynamics' Spot, are explicitly marked with "✖️" for license, indicating potential restrictions.

Limitations & Caveats

As a curated list, this repository does not provide any executable code or direct support. Users must individually assess the quality, compatibility, and licensing of each linked resource. The "updated gradually" note suggests that the list may not be exhaustive or entirely up-to-date.