Embodied Data Evaluation Engineer
Unitree Robotics is a Hangzhou-based public robotics company that develops and sells legged and humanoid robots for consumer, education, industrial, and research applications.
Investors
About Unitree Robotics
The brand Unitree Robotics is operated by Yushu Technology Co., Ltd. (宇树科技股份有限公司). Its current portfolio includes quadruped robots, humanoid robots, robotic arms, perception hardware, and associated support and developer resources.
Skills
Candidate Availability
Required and preferred rules are kept separate and reflect the wording in the original posting.
About the Role
You will design and implement distribution-modeling and processing systems for multimodal robot interaction data, and assess the value of large-scale embodied data. You will build automated data filtering, hard-example mining, and negative-sample generation mechanisms. You will establish evaluation systems and quantitative metrics linking data quality with model performance, and work with model and training groups to improve robot-operation generalization.
Requirements
- Proficiency in Python and mainstream deep-learning frameworks such as PyTorch and JAX
- Knowledge of statistical learning and machine learning
- Understanding of multimodal robot-data distributions
- Experience with large-scale data processing and distributed training
- Knowledge of robot kinematics and dynamics
Responsibilities
- Design and implement distribution-modeling and processing systems for multimodal robot interaction data
- Evaluate the value of large-scale embodied data
- Build automated data filtering, hard-example mining, and negative-sample generation mechanisms
- Establish evaluation systems and quantitative metrics for data quality and model performance
- Build data-model co-evolution processes with model and training groups to improve robot-operation generalization