Internship in Motion Tracking Algorithm Development - Remote

Overview

This internship focuses on developing an effective cross-embodiment motion tracking algorithm to enable humanoids to track human motions accurately while maintaining balance on various terrains.

In Short

• Creating a software workflow to transfer motions from SMPL-X to the Unitree G1 humanoid in simulators.
• Developing an algorithm for cross-embodiment motion tracking.
• Evaluating the performance of the tracking algorithm.
• Combining motion models with the tracking algorithm.
• Involvement in cutting-edge research in an international environment.
• Research experiences in both academia and industry.
• 6-month working contract.
• Encouragement of diverse applications to enhance collective achievements.

Requirements

• Enrolled student of a European university.
• Knowledge of PyTorch, deep learning, and neural networks.
• Good software development practices, including git and documentation.
• Basic knowledge of C++, reinforcement learning, and physical simulation.
• Basic knowledge of SMPL, SMPL-X, or character animation.
• Related experiences are a plus.
• Critical thinking and good communication skills.

Benefits

• Involvement in cutting-edge research.
• Research experiences in academia and industry.
• 6-month working contract.