Paper List:

Legged Robots:

Skill Discovery:

1. Emergent Real-World Robotic Skills via Unsupervised Off-Policy Reinforcement Learning [Paper]
2. DYNAMICS-AWARE UNSUPERVISED DISCOVERY OF SKILLS [Paper]
3. Learning Diverse Skills for Local Navigation under Multi-constraint Optimality [Paper]

Skill Learning/skill mixing:

1. Multi-expert learning of adaptive legged locomotion [Paper]
2. RSG: Fast Learning Adaptive Skills for Quadruped Robots by Skill Graph [Paper]
3. Imitate and Repurpose: Learning Reusable Robot Movement Skills From Human and Animal Behaviors [Paper] [website] [Imitation learning]
4. Walk These Ways: Tuning Robot Control for Generalization with Multiplicity of Behavior [Website]
5. Imitation Learning from MPC for Quadrupedal Multi-Gait Control [Paper]
6. Minimizing Energy Consumption Leads to the Emergence of Gaits in Legged Robots [Paper]
7. Agile But Safe: Learning Collision-Free High-Speed Legged Locomotion [Website]

Learning and MPC:

1. RL + Model-based Control: Using On-demand Optimal Control to Learn Versatile Legged Locomotion. [Paper]
2. Safe Importance Sampling in Model Predictive Path Integral Control. [Paper]
3. Leveraging Randomized Smoothing for Optimal Control of Nonsmooth Dynamical Systems [Paper]
4. First Order Approximation of Model Predictive Control Solutions for High Frequency Feedback [Paper]
5. A Hierarchical Scheme for Adapting Learned Quadruped Locomotion [Paper]
6. Learning Quadrupedal Locomotion over Challenging Terrain [Paper]
7. End-to-End Differentiable Physics for Learning and Control [Paper]
8. Optimization-Based Control for Dynamic Legged Robots [Paper]
9. DREAM TO CONTROL: LEARNING BEHAVIORS BY LATENT IMAGINATION [Paper]
10. Adaptive CLF-MPC With Application To Quadrupedal Robots [Paper]
11. Safe Importance Sampling in Model Predictive Path Integral Control [Paper]
12. Learning Robust and Agile Legged Locomotion Using Adversarial Motion Priors [Paper]
13. Learning a Single Policy for Diverse Behaviors on a Quadrupedal Robot using Scalable Motion Imitation [Paper]
14. OPT-Mimic: Imitation of Optimized Trajectories for Dynamic Quadruped Behaviors [Paper]
15. Gradient-Based Trajectory Optimization With Learned Dynamics [Paper]
16. DTC: Deep Tracking Control [Paper]
17. CoVO-MPC: Theoretical Analysis of Sampling-based MPC and Optimal Covariance Design [Website]
18. Reinforcement Learning for Versatile, Dynamic, and Robust Bipedal Locomotion Control [Paper]
19. ValueNetQP: Learned one-step optimal control for legged locomotion [Paper]
20. DeepGait: Planning and Control of Quadrupedal Gaits using Deep Reinforcement Learning [Paper]
21. Extreme Parkour with Legged Robots [Paper]
22. Barkour: Benchmarking Animal-level Agility with Quadruped Robots [Paper]
23. Temporal Difference Learning for Model Predictive Control [Paper] [Code] a. SIMPLIFYING MODEL-BASED RL: LEARNING REPRESENTATIONS, LATENT-SPACE MODELS, AND POLICIES WITH ONE OBJECTIVE [Paper]

Lagrangian Neural Newtork:

1. LNN [Medium] [code] [Paper]
2. DECONSTRUCTING THE INDUCTIVE BIASES OF HAMILTONIAN NEURAL NETWORKS [Paper]
3. Neural Networks with Physics-Informed Architectures and Constraints for Dynamical Systems Modeling [Paper]
4. Simplifying Hamiltonian and Lagrangian Neural Networks via Explicit Constraints [Paper] [code]
5. Lagrangian Model Based Reinforcement Learning [Paper]
6. Deep Lagrangian Networks for end-to-end learning of energy-based control for under-actuated systems [Paper]
7. Learning Contact Dynamics using Physically Structured Neural Networks [Paper] [code]
8. SYMPLECTIC ODE-NET: LEARNING HAMILTONIAN DYNAMICS WITH CONTROL [Paper]
9. Compositional Learning of Dynamical System Models Using Port-Hamiltonian Neural Networks [Paper]
10. Combining Physics and Deep Learning to learn Continuous-Time Dynamics Models [Paper] [code]
11. Extending Lagrangian and Hamiltonian Neural Networks with Differentiable Contact Models [Paper] [code]
12. ModLaNets: Learning Generalisable Dynamics via Modularity and Physical Inductive Bias [Paper]
13. Dissipative Hamiltonian Neural Networks: Learning Dissipative and Conservative Dynamics Separately [Paper]

Symmetries:

1. Symmetries of Lagrangians and of their equations of motion [Paper]

Rough Terrain Locomotion:

1. Adaptive Force-Based Control of Dynamic Legged Locomotion over Uneven Terrain [Paper]
2. Towards Legged Locomotion on Steep Planetary Terrain [Paper]
3. Learning Agile Locomotion and Adaptive Behaviors via RL-augmented MPC [Paper]
4. Visual-Locomotion: Learning to Walk on Complex Terrains with Vision [Paper]
5. DreamWaQ: Learning Robust Quadrupedal Locomotion With Implicit Terrain Imagination via Deep Reinforcement Learning [Paper]
6. Two-layer adaptive trajectory tracking controller for quadruped robots on slippery terrains [Paper]

Loco-Manipulation:

1. Hierarchical Adaptive Loco-manipulation Control for Quadruped Robots [Paper]
2. Contact Optimization for Non-Prehensile Loco-Manipulation via Hierarchical Model Predictive Control [Paper]
3. RoLoMa: robust loco-manipulation for quadruped robots with arms [Paper]

Transformer:

1. Decision Transformer: Reinforcement Learning via Sequence Modeling [Website]
2. Sim-to-Real Transfer for Quadrupedal Locomotion via Terrain Transformer [Paper]

Courses and related reads:

1. The Feynmann Lectures on Physics(V1), [Link]
2. Machine Learning and Physics: A Survey of Integrated Models [Paper]
3. The Cross-Entropy Method for Optimization [Paper]
4. WoLF: the Whole-body Locomotion Framework for Quadruped Robots [Github]
5. RL by David Silver [Link] [Code/Assignment]
6. Mastering Diverse Domains through World Models [Website]
7. Inverse Reinforcement Learning without Reinforcement Learning [Website]
8. Transfer Learning [Github]