Switch4EAI: Leveraging Console Game Platform for Benchmarking Robotic Athletics
🤖 AI Summary
Existing robot locomotion evaluation lacks standardized, real-world benchmarks enabling direct comparison with human motion. This work introduces a low-cost, easily deployable physical benchmark framework for embodied intelligence, pioneering the use of the commercial motion-based game *Just Dance* as a unified testbed. Leveraging motion capture and cross-modal retargeting, human dance sequences are precisely transferred to the Unitree G1 humanoid robot; full-body execution is achieved on real hardware via an open-source whole-body controller. The approach establishes a quantitative, reproducible standard for robot–human motion performance comparison. It successfully executes complete choreographies on physical hardware and enables end-to-end performance assessment. This work fills a critical gap in embodied intelligence benchmarks—particularly for dynamic whole-body control—and provides a scalable, real-world evaluation paradigm for developing and validating motion policies.
Application Category
📝 Abstract
Recent advances in whole-body robot control have enabled humanoid and legged robots to execute increasingly agile and coordinated movements. However, standardized benchmarks for evaluating robotic athletic performance in real-world settings and in direct comparison to humans remain scarce. We present Switch4EAI(Switch-for-Embodied-AI), a low-cost and easily deployable pipeline that leverages motion-sensing console games to evaluate whole-body robot control policies. Using Just Dance on the Nintendo Switch as a representative example, our system captures, reconstructs, and retargets in-game choreography for robotic execution. We validate the system on a Unitree G1 humanoid with an open-source whole-body controller, establishing a quantitative baseline for the robot's performance against a human player. In the paper, we discuss these results, which demonstrate the feasibility of using commercial games platform as physically grounded benchmarks and motivate future work to for benchmarking embodied AI.
Problem
Research questions and friction points this paper is trying to address.
Standardized benchmarks for robotic athletic performance evaluation are scarce
Leveraging motion-sensing console games to evaluate whole-body robot control
Establishing quantitative baseline for robot performance against human players
Innovation
Methods, ideas, or system contributions that make the work stand out.
Leveraging console game platform for benchmarking
Captures reconstructs retargets in-game choreography
Validates on humanoid with open-source controller
🔎 Similar Papers
No similar papers found.
