MAPO: Mixture of Experts with Alternating Policy Optimization for Universal Humanoid Motion Tracking
cs.RO
Submitted: 2026-11-02
Updated: 2026-11-02
Project page: https://mapo-corl2026.github.io/MAPO
Terminology
Sources
- Advancing Humanoid Locomotion: Mastering Challenging Terrains with Denoising World Model Learning
- Learning Humanoid Locomotion over Challenging Terrain
- Hiking in the Wild: A Scalable Perceptive Parkour Framework for Humanoids
- Learning Humanoid Standing-up Control across Diverse Postures
- Learning Getting-Up Policies for Real-World Humanoid Robots
- Expressive Whole-Body Control for Humanoid Robots
- BeyondMimic: From Motion Tracking to Versatile Humanoid Control via Guided Diffusion
- SONIC: Supersizing Motion Tracking for Natural Humanoid Whole-Body Control
- GMT: General Motion Tracking for Humanoid Whole-Body Control
- General Humanoid Whole-Body Control via Pretraining and Fast Adaptation
- Ray Interference: a Source of Plateaus in Deep Reinforcement Learning
- KungfuBot2: Learning Versatile Motion Skills for Humanoid Whole-Body Control
- Proximal Policy Optimization Algorithms
- Probabilistic Mixture-of-Experts for Efficient Deep Reinforcement Learning
- AMO: Adaptive Motion Optimization for Hyper-Dexterous Humanoid Whole-Body Control
- TWIST2: Scalable, Portable, and Holistic Humanoid Data Collection System
- Retargeting Matters: General Motion Retargeting for Humanoid Motion Tracking
- Isaac Lab: A GPU-Accelerated Simulation Framework for Multi-Modal Robot Learning
- Why M Heads are Better than One: Training a Diverse Ensemble of Deep Networks
Related papers
- Rethinking Actions: Continuity-Aware Priors for World-Model Planning
- MOSAIC: Modular Obstacle and Spatial-Aware Framework for Manipulation In Cluttered Scenes
- SAGAS: Semantic-Aware Graph-Assisted Stitching for Offline Temporal Logic Planning
- DexSeed: Expanding Human Hand-Object Interactions into Robot Demonstrations for Dexterous Policy Learning
- LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning
- CIALLO: Learning Shared Intent Representations for Cross-Embodiment Manipulation