Skip to content

Event

G1 achieves between-the-legs dribbling with deep vision reinforcement learning

1 reports · 1 sources

Guyuehome introduces a control method that enables the Unitree G1 humanoid robot to perform between-the-legs dribbling: it uses event-based rewards and asymmetric actor-critic training, allowing the depth camera to enter the observation loop directly during training, forming a G1 full-body dribbling policy. According to this introduction, when reproduced locally in MuJoCo using the official weights, the average over 12 episodes was 43.5 crossover dribbles with zero falls, and the results align with the official ones. This reproduction data comes from Guyuehome's description and is not independent verification.

Report timeline

  1. G1 Brunson: Humanoid Robot Between-the-Legs Dribbling Control Based on Deep Visual Reinforcement Learning

    G1 Brunson uses event-based rewards and asymmetric actor-critic training, bringing the depth camera directly into the observation loop during training to form a G1 whole-body dribbling policy. Reproducing the official weights locally in MuJoCo gave an average of 43.5 crossovers over 12 episodes with zero falls, matching the official results.

    Guyuehome ·

Related topics

Perception and localizationMotion controlRobot models and learningResearch papers

Interest trend

Current
3
Peak
7Nov 4 01:00
Past 24 hours
-50%

The trend compares participants observed throughout the same period, so its range may be smaller than the current interest count. Move or tap to inspect each hour; use the arrow keys on a keyboard.