Yuxiang Yang

I’m an AI Resident in Robotics at Google. My interest lies in the intersection of control and reinforcement learning with specific applications to robotics. Recently, my research is mostly involved with the legged locomotion of the Minitaur robot, which is particularly challenging due to its complex dynamics and real-time requirements. Before, I graduated from UC Berkeley on May, 2018 with a bachelor’s degree in Electric Engineering and Computer Science (EECS), and worked on legged locomotion in much smaller-scale millirobots with prof. Ronald Fearing.

 

Besides work, I enjoy cooking, running on Central Park and interacting with dogs. (real ones, not just robot dogs!)

 

Checkout my CV.

 

Publications

NoRML: No-Reward Meta Learning [ArXiV] [Website] [Code]

Yuxiang Yang, Ken Caluwaerts, Atil Iscen, Jie Tan, Chelsea Finn

Accepted for Oral Presentation at AAMAS 2019

We extended Model-Agnostic Meta-Learning (MAML) in the context of deep reinforcement learning. Our algorithm does not require ground-truth reward value during adaptation, is more responsive to changes in transition dynamics and sample efficient.

 

OpenRoACH: A Durable Open-Source Hexapedal Platform with Onboard Robot Operating System (ROS) [Arxiv] [Website]

Liyu Wang, Yuxiang Yang, Gustavo Correa, Konstantinos Karydis, Ronald S. Fearing

Accepted in ICRA 2019

We present a fully open-sourced, low-cost, legged robot that is capable of multi-terrain locomotion. The durable design allowed the robot to sustain several 24-hour burn-in tests. Its fully connected to ROS, allowing it to interact with a wide range of sensors.

 

Experiences

  • Google Brain Robotics, AI Resident, Jul 2018 - Jul 2020 (Expected)
  • UC Berkeley, Biomimetics Millisystems Lab, Feb 2017 - May 2018
  • UC Berkeley, Teaching Assistant, CS70, CS170, Aug 2016 - May 2018
  • Microsoft, Software Engineer Intern, May 2017 - Aug 2017
  • Google, Software Engineer Intern, May 2016 - Aug 2016

Education

UC Berkeley Bachelor of Science (May 2018)

Major: EECS, Minor: Mathematics, Highest Honor