CV
杨晨宇 Chenyu Yang
- Birth of Date: 7/23/1998
- Email: yangcyself@berkeley.edu
- Cell: US:+1 510-697-6884 CN:+86 13520183356
- Address: Shanghai Jiao Tong University No.800 Dongchuan Road, Shanghai 200240, China.
AREAS OF INTERESTS
- Robotics
EDUCATION
- School of electricity, Shanghai Jiao Tong University, Shanghai, China 09/2016 to 07/2019
- Major: Computer Science + Automation, Bachelor Engineering
- GPA 3.9/4.3 ranking 2/93
- Zhiyuan College, Shanghai Jiao Tong University 09/2016 to 07/2019
- Zhiyuan Honors Program of Engineering (Top 5%)
- Berkeley Global Access Discovery program 08/2019 to present
STANDARD TESTS
TOEFL: 103/120
GRE: 332/340 (91%)
SKILLS
Programming: C/C++, Python, MATLAB
- Robotic systems: ROS, v-rep
- ML systems: TensorFlow, Pytorch
- Modeling and analyzing control systems and design controllers
- Basic parallel programming with MPI, OpenMP and CUDA.
- Basic operations of building and managing a computing cluster.
HONORS & AWARDS
- 2017 Scholarship of Shanghai city 2%
- Mathematical Contest in Modeling Honorable Mention
LABORATORY EXPERIENCE
Dynamic Legged Manipulation of a Ball by Mini Cheetah Through Multi-Contact Optimization Dec. 2019 - Mar. 2020
Conducted as Intern project in Hybrid Robotics Lab, UC Berkeley. Supervised by Prof. Koushil Sreenath
- Implemented a controller and simulation environment of Mini Cheetah manipulating a ball.
- Achieved good performance than baseline.
- First author of the paper of the work published at IROS 2020
Human Driving Behavior Understanding with IRL Aug. 2019 - Feb. 2020
Conducted as Intern project in Mechanical Systems Control Lab, UC Berkeley. Supervised by Prof. Masayoshi Tomizuka
- Implemented a framework to learn rich and diverse human driving behavior with IRL.
- Achieved good prediction performance on real human driving behavior
- Analyzed difference of driver decision under different scenarios
- Fourth author of the paper of the work published at RAL 2020
- Helped in experiments of a paper published at ICRA2020
Advanced Computer Architecture Institute, SJTU
- Studied and carried out experiments about various machine learning tasks and hardware-based optimization.
Ministry of Education (MOE) Key Lab of Scientific and Engineering Computing, SJTU
- Studied and carried out experiments about reinforcement learning algorithms, legged robots, trajectory and locomotion planning algorithms.
Mechanical Systems Control Lab, UC Berkeley
- Carried out experiments about inverse reinforcement learning for self-driving cars.
SELECTED PROJECTS
- Inverse reinforcement learning for human driving behavior
- Use inverse reinforcement learning algorithms to learn the difference in driving behavior under different scenarios.
- Algorithm Designing for Trajectory Optimization
- Use reinforcement learning algorithms to guide the trajectory optimization algorithms of a hexapod.
- RL based Controller for Six-Legged Robot
- A C-class-conference paper
- The first writer for this work.
- Use reinforcement learning algorithms to train a controller for both robotic locomotion planning and trajectory planning.
- GitHub link: https://github.com/yangcyself/Hexpod_locomotion
- 2019 ASC Student Supercomputer Challenge
- Analyzed and Optimized the Community Earth System Model with SJTU HPC team as an inexperienced team member.
PUBLICATIONS
Dynamic Legged Manipulation of a Ball by Mini Cheetah Through Multi-Contact Optimization
Chenyu Yang, Bike Zhang, Jun Zeng, Ayush Agrawal, and Koushil Sreenath. Dynamic legged manipulation of a ball by mini cheetah through multi-contact optimization. InIEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), October 2020.
Efficient Sampling-Based Maximum Entropy Inverse Reinforcement Learning With Application to Autonomous Driving
Zheng. Wu, Liting. Sun, Wei. Zhan, Chenyu. Yang and Masayoshi Tomizuka, "Efficient Sampling-Based Maximum Entropy Inverse Reinforcement Learning With Application to Autonomous Driving," in IEEE Robotics and Automation Letters, vol. 5, no. 4, pp. 5355-5362, Oct. 2020