SoccerDiffusion: Toward Learning End-to-End Humanoid Robot Soccer from Gameplay Recordings
This work addresses the problem of enabling humanoid robots to perform soccer tasks autonomously, though it is incremental as it focuses on motion replication with limited high-level tactics.
The paper tackles learning end-to-end control policies for humanoid robot soccer from gameplay recordings, using a transformer-based diffusion model to predict joint trajectories from sensor inputs, with results showing replication of complex motions like walking and kicking in simulation and on physical robots.
This paper introduces SoccerDiffusion, a transformer-based diffusion model designed to learn end-to-end control policies for humanoid robot soccer directly from real-world gameplay recordings. Using data collected from RoboCup competitions, the model predicts joint command trajectories from multi-modal sensor inputs, including vision, proprioception, and game state. We employ a distillation technique to enable real-time inference on embedded platforms that reduces the multi-step diffusion process to a single step. Our results demonstrate the model's ability to replicate complex motion behaviors such as walking, kicking, and fall recovery both in simulation and on physical robots. Although high-level tactical behavior remains limited, this work provides a robust foundation for subsequent reinforcement learning or preference optimization methods. We release the dataset, pretrained models, and code under: https://bit-bots.github.io/SoccerDiffusion