Wenshan Wang

Project Scientist (2019 - current) - Carnegie Mellon University
Postdoctoral Fellow (2017 - 2019) - Carnegie Mellon University
Research Intern (2017) - Microsoft Research Asia
PhD / TA / RA (2009 - 2017) - Shanghai Jiao Tong University
Email: shanshan1370 [at] sina [dot] com
Major: Robotics
Curriculum Vitae | Research Statement

I have an obsessive interest in understanding intelligence. My research interests is in reinforcement learning and deep neural networks. I'd like to develop an end to end algorithm for robot perception, control and actuation based on reinforcement deep neural networks.


Project Experiences

Internship at Microsoft Research Asia
  • System framework and algorithm design.
  • Customer detection, tracking and localization using multiple cameras.
Internship at Microsoft Research Asia
  • Designing joint learning architecture for pixel-level prediction tasks such as optical flow, stereo and segmentation.
  • Developing robot navigation algorithm based on reinforcement learning and predictive model.
Internship at Pico China Ltd.
  • Developed applications of object recognition and voice recognition using open source deep learning frameworks.
  • Designed and developed a mobile platform with multiple sensors.
  • Developed an SLAM and navigation system based on Robot Operating System (ROS).
  • Other functionalities including face tracking and natural language processing.
Cooperated with Yaskawa Electrical Corporation
  • Designed and implemented the ubiquitous robotic system based on Robot Technology Middleware (RTM) and Robot Operating System (ROS).
  • Proposed a hierarchical task planning algorithm based on Markov Decision Process (MDP) model.
The Robotic Modularization Techniques (2013-2015)
The National High Technology Research and Development Program of China
  • Investigated the state of the art robotic software modularization techniques, and proposed a national standard on "Robotic Software Functional Component".
Standard platform league of Robocup soccer contest (2013-2014)
  • Team leader.
  • Multi-agent coordination, task planning based on hierarchical finite state machine.
Cooperated with Yaskawa Electrical Corporation
  • Proposed an algorithm for auto-creation of topological map with a probabilistic map build with the laser sensor.
  • Developed a robot navigation method for multi-layered environment based on A* algorithm.
  • Developed a graphical programming and simulation environment with Java3D for humanoid robot motion simulation and teaching.


  1. Rogerio Bonatti, Wenshan Wang, Cherie Ho, Aayush Ahuja, Mirko Gschwindt, Efe Camci, Erdal Kayacan, Sanjiban Choudhury, Sebastian Scherer, Autonomous aerial cinematography in unstructured environments with learned artistic decision‐making, Journal of Field Robotics, 2019.
  2. Mirko Gschwindt, Efe Camci, Rogerio Bonatti, Wenshan Wang, Erdal Kayacan, Sebastian Scherer, Can a Robot Become a Movie Director? Learning Artistic Principles for Aerial Cinematography, International Conference on Intelligent Robots and Systems (IROS), 2019.
  3. Xiangwei Wang, Daniel Maturana, Shichao Yang, Wenshan Wang, Qijun Chen and Sebastian Scherer, Improving Learning-based Ego-motion Estimation with Homomorphism-based Losses and Drift Correction, International Conference on Intelligent Robots and Systems (IROS), 2019.
  4. Rogerio Bonatti, Cherie Ho, Wenshan Wang, Sanjiban Choudhury, Sebastian Scherer, Towards a Robust Aerial Cinematography Platform: Localizing and Tracking Moving Targets in Unstructured Environments, International Conference on Intelligent Robots and Systems (IROS), 2019.
  5. Wenshan Wang, Aayush Ahuja, Yanfu Zhang, Rogerio Bonatti, Sebastian Scherer, Improved Generalization of Heading Direction Estimation for Aerial Filming Using Semi-Supervised Regression, International Conference on Robotics and Automation (ICRA), 2019.
  6. Rogerio Bonatti, Zhang Yanfu, Sanjiban Choudhury, Wenshan Wang and Sebastian Scherer, Autonomous drone cinematographer: Using artistic principles to create smooth, safe, occlusion-free trajectories for aerial filming, Proceedings of International Symposium on Experimental Robotics (ISER), 2018.
  7. Yanfu Zhang, Wenshan Wang, Rogerio Bonatti, Daniel Maturana and Sebastian Scherer, Xiaoxiao Zhu, Liyu Wang, Qiang Qiu, Qixin Cao, Integrating kinematics and environment context into deep inverse reinforcement learning for predicting off-road vehicle trajectories, Proceedings of Conference on Robot Learning (CoRL), 2018.
  8. Wenshan Wang, Xiaoxiao Zhu, Liyu Wang, Qiang Qiu, Qixin Cao, Ubiquitous Robotic Technology for Smart Manufacturing System, Computational Intelligence and Neuroscience, 2016.
  9. Wenshan Wang, Qixin Cao, Qiang Qiu, Gilbert Cheruiyot. Online learning of task models for ubiquitous robotic systems, International Conference on Intelligent Autonomous Systems. Springer, 2016, 1143-1154.
  10. Qixin Cao, Wenshan Wang, Xiaoxiao Zhu, Chuntao Leng, Study on Ubiquitous Robotic Systems for Smart Manufacturing Program, 2016 IEEE Workshop on Advanced Robotics and its Social Impacts (IASO2016), 164-169.
  11. Wenshan Wang, Qixin Cao, Modeling for Robot Task Planning based on Light-weighted Markov Decision Process, Journal of Huazhong University of Science and Technology, 2015, 43(S1): 58-61.
  12. Wenshan Wang, Qixin Cao, Xiaoxiao Zhu, Shuang Liang, A Framework for Intelligent Service Environments Based on Middleware and General Purpose Task Planner, 2015 International Conference on Intelligent Environments (IE), IEEE, 2015, 184-187.
  13. Wenshan Wang, Qixin Cao, Xiaoxiao Zhu, Masaru Adachi. An automatic switching approach of robotic components for improving robot localization reliability in complicated environment. Industrial Robot: An International Journal, 2014, 41(2): 135-144.
  14. Wenshan Wang, Qixin Cao, Chengcheng Deng, Zhong Liu, Auto-Creation and Navigation of the Multi-area Topological Map for 3D Large-Scale Environment, LSMS/ICSEE 2010, Springer-Verlag Berlin Heidelberg, 2010, 307-315.

Honors and Awards

For more information, please contact me by email