# Bi-Level Motion Imitation for Humanoid Robots

arXiv: https://arxiv.org/abs/2410.01968 · HTML: https://mlsanity.com/paper/2410.01968/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Real-world robot offline RL, one of 13 directions in Offline RL and imitation learning from data.
- First posted on arXiv 2 Oct 2024.

## Placed in

- Offline RL and imitation learning from data / [Real-world robot offline RL](https://mlsanity.com/offlinerl/real-world-robot-offline-rl/)

## Built on by (most cited first, 4 of 4)

- [ReActor: Reinforcement Learning for Physics-Aware Motion Retargeting](https://mlsanity.com/paper/2605.06593/) (2026)
- [BarlowWalk: Self-supervised Representation Learning for Legged Robot Terrain-adaptive Locomotion](https://mlsanity.com/paper/2508.00939/) (2025)
- [Shooting for Contact: Contact-Implicit Multiple Shooting for Dynamic Motion Retargeting](https://mlsanity.com/paper/2608.03116/) (2026)
- [DreamHMR: From Video Generation to Human Motion Modeling via Structural Focusing](https://mlsanity.com/paper/2610.20124/) (2026)

## Abstract (arXiv)

Imitation learning from human motion capture (MoCap) data provides a promising way to train humanoid robots. However, due to differences in morphology, such as varying degrees of joint freedom and force limits, exact replication of human behaviors may not be feasible for humanoid robots. Consequently, incorporating physically infeasible MoCap data in training datasets can adversely affect the performance of the robot policy. To address this issue, we propose a bi-level optimization-based imitation learning framework that alternates between optimizing both the robot policy and the target MoCap data. Specifically, we first develop a generative latent dynamics model using a novel self-consistent auto-encoder, which learns sparse and structured motion representations while capturing desired motion patterns in the dataset. The dynamics model is then utilized to generate reference motions while the latent representation regularizes the bi-level motion imitation process. Simulations conducted with a realistic model of a humanoid robot demonstrate that our method enhances the robot policy by modifying reference motions to be physically consistent.