Job Number: P22INT-23

The focus of the research is to extend machine learning techniques (i.e. behavior cloning, imitation, reinforcement) to enable multi-fingered robot hand to dexterously manipulate objects using vision and tactile sensor data. The candidate is expected to utilize both simulation and the real hardware to implement the result on the learnt policy.

San Jose, CA

Key Responsibilities:

• Develop algorithms to learn policies to manipulate an object on the multi-fingered robot hand.
• Explore various machine learning techniques to utilize available data on simulated data and the real data.
• Implement and validate the developed algorithms in simulation and on hardware.

Minimum Qualifications:

• Ph.D. or highly qualified M.S. candidate in computer science, robotics, or related field.
• Experience in deep learning and other machine learning methods.
• Experience in robotics manipulation, control, path planning.
• Good programming skills in Python and C++.
• Experience with Robot Operating System (ROS).

Bonus Qualifications:

• Experience with sim-to-real approaches.
• Experience in multi-fingered robotics hand, dexterous manipulation.
• Experience in Deep Reinforcement Learning, Imitation Learning, Model Predictive Control.
• Experience with PyTorch or TensorFlow.

Duration: 3 ~ 4 months

How to apply

Candidates must have the legal right to work in the U.S.A. Please add Cover Letter and CV in the same document