mlsanityData to 13 Oct 2026

From Universal Humanoid Control to Automatic Physically Valid Character Creation

arXiv 2206.09286 ↗First posted 18 Jun 2022cs.GR cs.CVcited by 11

Not on the map. Why?

Abstract (arXiv)

Automatically designing virtual humans and humanoids holds great potential in aiding the character creation process in games, movies, and robots. In some cases, a character creator may wish to design a humanoid body customized for certain motions such as karate kicks and parkour jumps. In this work, we propose a humanoid design framework to automatically generate physically valid humanoid bodies conditioned on sequence(s) of pre-specified human motions. First, we learn a generalized humanoid controller trained on a large-scale human motion dataset that features diverse human motion and body shapes. Second, we use a design-and-control framework to optimize a humanoid's physical attributes to find body designs that can better imitate the pre-specified human motion sequence(s). Leveraging the pre-trained humanoid controller and physics simulation as guidance, our method is able to discover new humanoid designs that are customized to perform pre-specified human motions.

Key facts. Cited by 11 papers in the map (data through 13 Oct 2026). Builds on 14 of its 27 references that are in the map. First posted on arXiv 18 Jun 2022.

Builds on

14 of 27 references.

References · 14

  1. 2017Proximal Policy Optimization Algorithmscited by 14,592
  2. 2016OpenAI Gymcited by 1,694
  3. 2019AMASS: Archive of Motion Capture as Surface Shapescited by 1,213
  4. 2021SimPoE: Simulated Character Control for 3D Human Pose Estimationcited by 113
  5. 2021D-Grasp: Physically Plausible Dynamic Grasp Synthesis for Hand-Object Interactionscited by 110
  6. 2021Dynamics-Regulated Kinematic Policy for Egocentric Pose Estimationcited by 101
  7. 2020Residual Force Control for Agile Human Behavior Imitation and Extended Motion Synthesiscited by 63
  8. 2018Reinforcement Learning for Improving Agent Designcited by 55
  9. 2018Jointly Learning to Construct and Control Agents using Deep Reinforcement Learningcited by 51
  10. 2020UniCon: Universal Neural Controller For Physics-based Character Motioncited by 39
  11. 2019Neural Graph Evolution: Towards Efficient Automatic Robot Designcited by 36
  12. 2021Transform2Act: Learning a Transform-and-Control Policy for Efficient Agent Designcited by 35
2 more
  1. 2019Data-efficient Co-Adaptation of Morphology and Behaviour with Deep Reinforcement Learningcited by 34
  2. 2021A Survey on Deep Learning for Skeleton-Based Human Animationcited by 27

Built on by

Cited by year (2026 to 13 Oct)

Citing papers per year: 2023: 1; 2024: 1; 2025: 5; 2026: 4.

11 papers.

  1. 2023Perpetual Humanoid Control for Real-time Simulated Avatarscited by 201
  2. 2024MaskedMimic: Unified Physics-Based Character Control Through Masked Motion Inpaintingcited by 103
  3. 2025ModSkill: Physical Character Skill Modularizationcited by 7
  4. 2025Zero-Shot Human-Object Interaction Synthesis with Multimodal Priorscited by 7
  5. 2025Generating Physically Realistic and Directable Human Motions from Multi-Modal Inputscited by 6
  6. 2025SimGenHOI: Physically Realistic Whole-Body Humanoid-Object Interaction via Generative Modeling and Reinforcement Learningcited by 3
  7. 2025Endowing GPT-4 with a Humanoid Body: Building the Bridge Between Off-the-Shelf VLMs and the Physical Worldcited by 2
  8. 2026Iterative Closed-Loop Motion Synthesis for Scaling the Capabilities of Humanoid Controlcited by 1
  9. 2026BRIDGE: An Open-Source Humanoid Platform via Morphology-Control Co-Design for Physical AIcited by 1
  10. 2026Policy-as-Data: Learning Generalizable HOI Diffusion Models from Simulated Physicscited by 0
  11. 2026Progressively Learning Heterogeneous Skills in a Unified Latent Spacecited by 0